Design of temporary pad foundation

[Zambo]

I have a question about the design of a pad foundation. We had an SI carried out which included borehole log, SPT results and lab soil tests. The SI was professionally done and then this was given to a structural engineer to carry out the design of a temporary pad foundation for a tower crane (tower crane max capacity 20T and 220m high so not a small job)

The structural engineer averaged the N values over a depth equal to the width of the foundation and then proceeded with his design. The problem is that the N values ranged from 0-10 giving an average of 6. Because the upper soils had an N value of 0 it was decided to carry out a plate bearing test to check. This plate bearing test resulted in a much lower bearing capacity than using the average of N values over a depth equal to the foundation width.

Finally instead of using the N values and averaging them we just excavated a bit deeper, used a plate bearing test and designed and constructed a foundation 2.5 times the original plan area.

Is it the case that with low N values of 0-2 they cannot be averaged in the normal way? Or could we have proceeded with the averaged value and not carried out a plate bearing test?

 

[ishvaaag]

I think it was a sound decision to go deeper. So low N means also the risk of localized failure is bigger on the variability of the strength, and with so tall structure this might mean significant risk.

 

[TDAA]

I think that it is more likely that averaging the values is inappropriate at those low values. Sure, if they are 20-30, then 25 seem reasonable.

The risk of localized failure is what I would be concerned about. When you start getting to that low of a count, you are likely dealing with mud at various points.

When looking at whether or not it is wise to average, or perhaps more appropriately how to average, the results, it is best to look at stratification, and average similar results only.

 

[Zambo]

ishvaag and TDAA,

thanks for your responses. The crane is tied-back to the permanent structure which helps with the loading. Is there any rule of thumb about when the N values can be averaged or is it a matter of using experience?

 

[oneintheeye]

average them over similar strata i.e do not average rock against mud. You can also increase by a factor dependent of overburden, this can be up to 2 at lower depths. There are charts available in standard text books.