Soldier Beam Design Methodology 2

[jdgengineer]

I've been working on updating my soldier beam spreadsheet and have a few questions. I had previously posted a version that used the "simplified" approach but have now retailored it to the "conventional" approach shown in USS Sheet Pile Design Manuel (modified for soldier beams) and Caltrans Trenching & Shoring Manual. Attached is print out of input / output from a typical temporary shoring condition. This would be used for a relatively shallow excavation for a basement in a single-family residential neighborhood. The adjacent structure in this example is far enough away to avoid surcharging the wall.

The values used in the attached design include:

Surcharge - 72 PSF min (or actual surcharge if higher). 72 PSF is recommended by Caltrans Trenching & Shoring Manual
Active / Passive - as prescribed by geotech
Factor of Safety - 1.3 (recommended by Caltrans Trenching & Shoring Manual)
Neglected Depth - 1 ft (typically recommended by geotech)
Starting Passive Pressure - Passive Pressure x Neglected Depth to provide trapezoidal distribution
Effective Passive Factor - 2-3 as recommended by geotech
Effective Active Pressure - 1x pier diameter below dredge line
Cover on soldier beam - 1" min for temporary condition

These values often tend to produce 1.5:1 embed to retained height.

Some questions...

1) most of the examples for soldier beam calculations I've seen apply the active pressure the entire length of the pier including depth below the dredge line. I believe the thought is that to engage passive you need to have fairly significant movement so you will have active on the backside of the pier. Some calculations I've seen don't do this (Retain Pro module as an example) and instead utilize a form of the IBC cantilevered pole equation. Site soils will likely play into the appropriate approach, but what do you see as typical approach? When do you use one vs the other?

2) The cantilevered pole equation appear to have stemmed from Pole Building Design by Donald Patterson. In this assumptions, it looks like the moment arm couple to provide fixity for the pier has a longer lever than it assume with this approach which results in slightly shallower embedments (say 10-20%). Have you tried to use a similar philosophy for these embedments?

3) if you apply the active times the full depth of the pier, do you use the same effective width as used for the passive pressure or do you only use the width of the pier? Caltrans Trenching & Shoring Manual uses the same for both active and passive and so does other references (Army Corps I believe does as well based on f factor to convert sheet pile to soldier beam) but Civiltechs software manual appears to only use the pier diameter (I don't have the software just read the manual). Other shoring drawings I've seen from engineers also only use the pier diameter. The geotech I discussed this with didn't seem to agree with the concept of not using the same effective width for both active and passive.

4) Do you typically apply minimum surcharge for these conditions? The retaining side does not have traffic but is neighboring property where we cannot dictate to them not to do certain things in their yard during construction. Most similar shoring calculations in our area do not include any surcharge except where undermining existing structures.

5) What factor of safety do you use? I've seen other people use values as low as 1. 1.25-1.3 seems to be commonly recommended for temporary. Some geotechs include factors of safety in the passive values they give us. We try and parse our an ultimate gross passive value (some also provide net instead) for our design purposes.

6) What cover do you use on soldier beam for temporary conditions? Smaller cover allows us to provide deeper soldier beams with same size hole and therefore limit deflections a little with same steel tonnage.

7) Most geotechs require that we neglect the top 1'-0" in our design. While I understand the concept, it seems since we are so far below the natural ground surface this may be a bit conservative. Thoughts?

 

[DaveAtkins]

SPW-911 gives the factor of safety against failure of the toe. If you input the exact required embedment, as calculated by Teng, SPW-911 says the FS = 1.0. If you increase the embedment 20%, SPW-911 says the FS is approximately 1.5.

DaveAtkins

 

[jdgengineer]

Interesting, thanks for the input here guys. I appreciate all the time and effort in the responses. Especially your last post Pe,Inc. Really above and beyond.

It looks like I am in the minority here, but I remain unconvinced. It's ok, as you all have said there is no one correct way and the soil properties are not an exact science regardless so sharpening the pencil too much is somewhat pointless.

I still firmly believe the conventional method is the traditional way of assuming the soil pressure (assuming rigid pile which is flawed on its own). As this approach is quite onerous the simplified approach was developed to make the process much easier to calculate by hand. The simplification introduces some errors that need to be corrected for.

It is my understanding that the simplified approach will always have smaller embedments than the conventional approach. This is why I think it then needs to be corrected to embedments calculated by the conventional method. However, based on PE,Inc analysis and my spreadsheet comparison it seems like the correlation factor is likely closer to 1.05-1.1 than 1.2. Therefore, the 1.2 factor is a bit conservative (it appears) to be used against the depth and then again as a factor of safety. However, I still believe applying just the 1.2 factor against the D calculated using Teng's approach does not provide the same factor of safety as you would have by applying the 1.2 against the conventional method. Perhaps, the differences are small enough and can be chalked up to falling within the margin of error for design soil properties. I'm not a geo so I'm not an expert on that but I'll defer to you experts.

Thanks again for all the help. I appreciate the discussion. I posted initially on the structural forum and didn't get a single reply. Dang structural engineers get scared when they see dirt instead of steel

 

[PEinc]

So, tell me what the "errors" are with the simplified system. What is worse - small "errors" in a simple method or elaborate, theoretical assumptions in the conventional method? Essentially, the conventional method nets out the active and passive pressures to give a combined diagram. Then, crazy assumptions are made to try equalizing the horizontal forces. However, passive resistance is not an applied force. It only builds up enough (hopefully) to resist the driving forces. For a cantilevered sheeting wall, the total passive resistance must always be greater than the total driving forces because their respective moment arms are much different. Passive arm is much shorter. Therefore more resistance is needed. IMHO, trying to equalize the horizontal force is unnecessary for a cantilevered wall. The wall cannot move backwards from a passive resistance that is greater than the driving force.

Also, I have my own problems with SPW-911, especially for soldier beams; but that's a different thread.

http://www.PeirceEngineering.com

 

[jdgengineer]

PEInc you understand that the simplified method is a simplified version of the conventional method? Any errors in the theoretical assumptions of the conventional method are also errors in the simplified method. The main difference between the two is that the simplified method is the conventional method but stops the analysis at the pivot point and introduces a fictitious reaction at depth Do that is intended to represent the passive pressure below the pivot point. As the location of this reaction does not occur at the pivot point as the simplified method assumes you need to calibrate the depth by multiplying Do x 1.2 Aside from this assumptions of both methods are the same. The main difference is that the conventional method directly determines the length of the embedment required below the pivot point where the simplified method does not and instead estimates this distance as "1.2 x Do". Both methods assume triangular active / passive pressure loading and rigid poles. Both of these assumptions introduce error but the simplified method introduces an additional error of estimating the length of embedment below the pivot point as "1.2 x Do".

Now I know you don't agree that this is what the simplified method is doing so I'm not entirely sure how else to explain it. But that is what it is doing.

Page 20 of the USS Sheet Pile Design Manual -"A cantilevered sheet pile wall may be designed in accordance with the principles and assumptions just discussed or by an approximate method based on further simplifying assumptions shown in Figure 15" Pay close attention to Figure 15. You'll note that 1.2 Do is drawn to the exact same depth as the "D" determined from the conventional method

Page 23 of the USS Sheet Pile Design Manual - "The passive resistances are simplified by assuming a right triangular pressure on the left side of the piling and by substitution of a concentrated force C for the net passive resistance on the right side of the piling This method results in some error but saves greatly in the computations. The distance, Do, must satisfy both the requirements of equilibrium. The calculated value of Do should be increased by 20 to 40 percent to get the total design depth of penetration

 

[PEinc]

My last comments:

In my opinion, the entire focus of the conventional method for cantilevered sheeting design is to develop force equilibrium in the horizontal direction after determining a depth of embedment for moment equilibrium. However, a cantilevered sheeting wall is not a free body diagram. The wall cannot push backward as a result of the greater passive "pressure." If the wall tried to push back the soil behind the wall would become passive. Would this now much greater, full height force now try to push the wall forward? No. If properly designed to resist overturning, the wall cannot move forward just because the total passive resistance is greater than the driving force; and the wall cannot move backwards. The simplified method addresses moment equilibrium (overturning) and then puts a safety factor on the embedment and therefore on the overturning. If you are worried about horizontal force equilibrium, why don't you find D for moment equilibrium, increase it by at least 20%, and then divide the new total passive resistance by the total driving forces (active and surcharges) to get a safety factor against sliding?

For over fifty years, people have been tweaking Teng's simplified method, beyond what Teng's clearly wrote, in order to make it more conservative. It's too bad we can't ask Teng to settle this.

http://www.PeirceEngineering.com

 

[jdgengineer]

Thanks for your continued input. I've appreciated the dialog and the different views. While I think we still do not agree I think it was a good discussion and I respect your opinion and vast experience.

I think what you described about the wall pushing back below the pivot point is infact the baseline modeling assumption for these piers. The simplified method uses the same assumption just simplifies the calculation process. If you read Teng's book or USS Sheet Pile both state this as the case. And yes, the passive pressure is assumed to flip to the other side below the dredge line. This provides you a "moment couple" to provide the fixity. It's my understanding that this is how the system would work and if you run a PY analysis I think you will see that it has the same conclusion.

I agree it would be nice to have Teng able to provide clarity to settle this. Having not practiced as long as you, I haven't seen his method be refined over the years with more conservative approaches. Had I witnessed that I my viewpoint may better align with yours.

For now I think we will just have to agree to disagree.