Questions regarding using average values of SPT-N values 2

[sworddark90]

Hello , i have been given a task to record spt values in a 5 layers , each layer is 1 meter thick. My question is , in reference to the attached boring log , is it good practice to take the average values of N between layers or to take the minimum value (For example 1-2.0m depth , do i take the N value at 1.5m depth or take the average between 1.0 and 2.0m depths). thanks in advance

 

[SlideRuleEra]

IMHO, depends on use of the data.
As a former bridge contractor, when driving piling I want to see all the N values (neither average nor minimum) to get the best idea of driving conditions.

If you have to provide just one value, perhaps it could be the average with plus or minus to show variation. For example: N = 20+4

 

[sworddark90]

Shallow foundation (3.5mx3.5m)

 

[EireChch]

Recording them in an arbitrary increment of 1m thick layers is not good practice I dont think.

Plot them graphically against depth. Then create a design line. From a quick look it seems you have a medium dense layer from 0-2m depth, loose to 4m depth, medium dense to 6.5m where very dense is encountered.

 

[sworddark90]

I have been asked to find the settlement using Schmertmann's method , which requires layer divisions.

 

[EireChch]

Well if thats what you are doing why dont you just give one layer for each SPT increment.

No one can critsice your averages then

 

[sworddark90]

Yea could do that , but lets say for the case of 2.5x2.5m square footing , where the stresses are taken to 2B from the foundation level , hence 2(2.5)= 5m , Hence in this case where significant stresses are within 2B , how can we go about N values if I am limited to 5 layers? do I take the minimum value of N between layers or average ? How much inaccurate will the results be if choosing either procedure? Which one more suitable you think? Taking minimum values are in the conservative side while average values are more accurate (i think?) Thanks for your cooperation.

 

[EireChch]

If I had to do you exact task I would take average. Lowest is too conservative.

Edit, actually I wouldnt just take average. Any time I get data i plot it on a graph. 99% of the time its several BHs so it has to be plotted to get an idea of the data. For your case you have one BH. I would probably take what I call lower bound average (i.e take average and knock a little off).

For your layer of 0-2m. I would take 16+24+16+23 = 79 /4 = 20. Id probably take 18.

Someone could also argue that you are basing all your calcs on one BH. That's fine if your pad is going at that exact location. If you have one BH that is covering a larger area then there could be variation that you have not investigated. Then maybe you should use lowest value.

It all comes down to engineering judgement, and even that varies a lot between engineers

Hence why I said do a layer for each SPT, removes all possible criticism.

Side note, I also think 1m thick layers is too thick. use 0.5m thick layers.