Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
spt correlations to su 1
- Thread starter ngreece
- Start date
Pretty obvious if you have the figure, you can determine an equation - in Singapore, it is common to use Su = 5N (but couldn't find out if the N is corrected or not . . . will try to find out). I used to make my own correlations in given locales - site specific is always better than the "general" ones given their very fuzzy nature . . .
Just be careful about what's being correlated with what. Not sure about Sowers, but the Terzaghi and Peck correlation is N (not adjusted for any of the things we now adjust for) and unconfined compressive strength, qu, not Su.
There exists a little-known correlation of N with field vane shear, which has some advantages (primarily less concern about sample disturbance), and which I believe I have referenced on another thread recently. The disturbance may have something to do with the fact that, for a given N, Sowers gives higher qu with more plastic material.
DRG
There exists a little-known correlation of N with field vane shear, which has some advantages (primarily less concern about sample disturbance), and which I believe I have referenced on another thread recently. The disturbance may have something to do with the fact that, for a given N, Sowers gives higher qu with more plastic material.
DRG
-
1
- #5
The correlation:
Su/Pa = 0.06*N
Is given by Terzaghi & Peck, Pa is atmospheric pressure expressed in the unit you wish to use, and, according to Kulhawy & Mayne, 1990, it has been developed from unnconfined compression tests.
As dgillette implied, originally it was qu =12N in kPa (as cited in Bowles' handbook
Itoo have always wondered wether the N values are corrected or not but, unless specified as N1(60), or N1(70), I'll assume they are not.
The Stroud & Butler correlation is so similar to the one which relates constranend modulus to NSPT and PI that it makes me suspicious, even because it would yield pretty low Su values in low plasticity clays.
The Terzaghi correlation I often feel may underestimate strenght, I have no data to support it though, anyway it's located in a cautious region in between the various suggested correlations.
Su/Pa = 0.06*N
Is given by Terzaghi & Peck, Pa is atmospheric pressure expressed in the unit you wish to use, and, according to Kulhawy & Mayne, 1990, it has been developed from unnconfined compression tests.
As dgillette implied, originally it was qu =12N in kPa (as cited in Bowles' handbook
Itoo have always wondered wether the N values are corrected or not but, unless specified as N1(60), or N1(70), I'll assume they are not.
The Stroud & Butler correlation is so similar to the one which relates constranend modulus to NSPT and PI that it makes me suspicious, even because it would yield pretty low Su values in low plasticity clays.
The Terzaghi correlation I often feel may underestimate strenght, I have no data to support it though, anyway it's located in a cautious region in between the various suggested correlations.
Hi McCoy. The correlation referenced to VST is
Su = 600 lb/ft^2 + 140 N +/- 500 lb/ft^2.
(I fit that equation to plotted data from a USBR research report.) Those tests were done in the early 1960s, and were almost certainly not adjusted for hammer energy or anything. Probably a doughnut hammer, but that's far from certain. I knew the guy who did the tests, but I can't ask him because he's dead now (and, coincidentally, he's buried at one of the test sites, which was at that time a new cemetery being built). The VST data have not been adjusted a la Bjerrum or Aas, LaCasse, et al.
I don't know how to find the Stroud reference (hint!), but in 1975 they probably were not making quantitative adjustments.
DRG
Su = 600 lb/ft^2 + 140 N +/- 500 lb/ft^2.
(I fit that equation to plotted data from a USBR research report.) Those tests were done in the early 1960s, and were almost certainly not adjusted for hammer energy or anything. Probably a doughnut hammer, but that's far from certain. I knew the guy who did the tests, but I can't ask him because he's dead now (and, coincidentally, he's buried at one of the test sites, which was at that time a new cemetery being built). The VST data have not been adjusted a la Bjerrum or Aas, LaCasse, et al.
I don't know how to find the Stroud reference (hint!), but in 1975 they probably were not making quantitative adjustments.
DRG
Dave - Su = 5N (kPa) - as indicated by Nick Shirlaw (Golders) - seems to be the "norm" for Singapore clays. In Toronto, we used the typcial Su ~ 6.25N (as given in most books for very soft, soft, firm, stiff, very stiff). Allowable bearing pressures were typically taken as qall = 10N (makes sense qu ~ 6Su; qu ~ 6x6.25N ~ 40N; SF=3; qall ~ 13N; use 10 for ballpark).
dgillette,
I converted your VST corelation in atmospheres:
Su = 0.28+0.066N +/- 0.24
An N value of 10 would result in a Su = 94 +/- 24 kPa, versus the 60 kPa Su as per Terzaghi's correlation, 50 to 62.5 Su of BigH's singaporean correlation.
Your correlation is apparently more suited to the kind of clays I usually come across over here, and I love the presence of the error value.
I often feel the Terzaghi correlation underestimates Su in medium and low plasticiy silty clays.
Is the source for the USBR correlation available anywhere?
I like the story of the guy who did the tests and now has a pretty direct feel of the very ground he tested.
Stroud's corelation: I'm suspicious in the sense the cited correlation would appear to be the one for constrained modulus, not Su, the reference is Stroud 1974. ESOPT:"The SPT in insensitive clays and soft rocks"
and Stroud & Butler 1975 : THE SPT and engineering properties of glacial materials.
In a few words, Stroud 1974 and Stroud and Butler 1975 published correlations for the NSPT versus the constrained modulus, not for the NSPT versus Su, as far as I know.
I did not look the above articles up though, so I might not be perfectly accurate in my statement.
I converted your VST corelation in atmospheres:
Su = 0.28+0.066N +/- 0.24
An N value of 10 would result in a Su = 94 +/- 24 kPa, versus the 60 kPa Su as per Terzaghi's correlation, 50 to 62.5 Su of BigH's singaporean correlation.
Your correlation is apparently more suited to the kind of clays I usually come across over here, and I love the presence of the error value.
I often feel the Terzaghi correlation underestimates Su in medium and low plasticiy silty clays.
Is the source for the USBR correlation available anywhere?
I like the story of the guy who did the tests and now has a pretty direct feel of the very ground he tested.
Stroud's corelation: I'm suspicious in the sense the cited correlation would appear to be the one for constrained modulus, not Su, the reference is Stroud 1974. ESOPT:"The SPT in insensitive clays and soft rocks"
and Stroud & Butler 1975 : THE SPT and engineering properties of glacial materials.
In a few words, Stroud 1974 and Stroud and Butler 1975 published correlations for the NSPT versus the constrained modulus, not for the NSPT versus Su, as far as I know.
I did not look the above articles up though, so I might not be perfectly accurate in my statement.
What are you using the correlation for?
Is it a critical project?
Then I would suggest that this approach is not appropriate and that you should attempt to recover undisturbed samples and do triaxial/shear box testing.
Or,
Perform oedometer testing and calculate OCR values and use a standard overburden pressure vs su assumption.
Or,
Estimate su from field vane testing which records stress vs strain throughout the duration of the test.
The SPT test is not an undrained test. The rate of loading and the input energy is highly variable and as such the test is not reliable for anything other than pre feasibility level guesstimates of su and definitely not suitable for design purposes.
Is it a critical project?
Then I would suggest that this approach is not appropriate and that you should attempt to recover undisturbed samples and do triaxial/shear box testing.
Or,
Perform oedometer testing and calculate OCR values and use a standard overburden pressure vs su assumption.
Or,
Estimate su from field vane testing which records stress vs strain throughout the duration of the test.
The SPT test is not an undrained test. The rate of loading and the input energy is highly variable and as such the test is not reliable for anything other than pre feasibility level guesstimates of su and definitely not suitable for design purposes.
![[peace]](/data/assets/smilies/peace.gif "[peace] [peace]")
Finally got around to scanning it last night when scanning other things.![[cheers]](/data/assets/smilies/cheers.gif "[cheers] [cheers]")
the best way to estimate undrained shear strength from SPT is to jam a pocket penetrometer into the recovered SPT sample.
I completly agree that correlations between blow counts and undrained shear strength are iffy.
f-d
¡papá gordo ain’t no madre flaca!
I completly agree that correlations between blow counts and undrained shear strength are iffy.
f-d
¡papá gordo ain’t no madre flaca!
1. In addition to Delijosi post about the relation of Stroud (1975) some more information on values of f1 (all in kPa):
f1?4.5 for PI>30%
f1?5 for PI=25%
f1?7 for PI=15%
2. According to Sowers in kPa (I found the relation from graph of NAVFAC DM7.01 with the adequate conversion):
f1=3.6 for clays of low plasticity and clayey silts
f1=7.2 for clays of medium plasticity (close to Terzaghi value)
f1=12 for clays of high plasticity.
3. As one can note, those two approaches contradict. With increasing plasticity, the first gives lower Su and the second higher Su for the same blow counts!
4. Consequently those relations should be used with caution.
5. Don't forget that SPT is a quite good tool for sands but not for clays.
f1?4.5 for PI>30%
f1?5 for PI=25%
f1?7 for PI=15%
2. According to Sowers in kPa (I found the relation from graph of NAVFAC DM7.01 with the adequate conversion):
f1=3.6 for clays of low plasticity and clayey silts
f1=7.2 for clays of medium plasticity (close to Terzaghi value)
f1=12 for clays of high plasticity.
3. As one can note, those two approaches contradict. With increasing plasticity, the first gives lower Su and the second higher Su for the same blow counts!
4. Consequently those relations should be used with caution.
5. Don't forget that SPT is a quite good tool for sands but not for clays.
There isn't necessarily a contradiction there, and there may be an explanation.
Go back and look at the strength tests that the SPTs were correlated with. In the case of T&P, and I believe also Sowers, the reference test was the unconfined compressive strength. Less-plastic materials suffer more disturbance in sampling and handling, which may explain why they (at least on average) give lower qu for a given N. Regarding Stroud and Butler, Delijosi wrote "SPT N value was referenced to lab Su," but was not specific. Assuming that was something other than unconfined, the effects of disturbance are compensated to some extent; hence, there could be less of a trend or an opposite trend in Su vs PI, for a given N.
How good were the fits? [I haven't seen the actual data for either T&P or Sowers (or Stroud).] Did they control for all the relevant variables that would create apparent trends where none exist? For example, did all of their high-PI data come from a particular stratum that happened to be at a greater depth or had higher Ns? One correlation I remember, but can't find, shows the slope of Su vs N decreasing with increasing N. Was that effect involved? Were the same hammers used at all the sites, or did some get tested with doughnut hammers while others had CME auto hammers? Does sensitivity affect it, and can that sensitivity be linked with PI in such a way as to explain the trend? Were enough different sites and materials included to show that the trends are valid in general and not just between site A and site B?
Considering the mechanism by which clayey soil resists the penetration of split-barrel sampler, is there any theoretical reason why a more (or less) plastic clay with a given Su should provide greater penetration resistance?
BTW - where was Stroud and Butler 1975 published?
Go back and look at the strength tests that the SPTs were correlated with. In the case of T&P, and I believe also Sowers, the reference test was the unconfined compressive strength. Less-plastic materials suffer more disturbance in sampling and handling, which may explain why they (at least on average) give lower qu for a given N. Regarding Stroud and Butler, Delijosi wrote "SPT N value was referenced to lab Su," but was not specific. Assuming that was something other than unconfined, the effects of disturbance are compensated to some extent; hence, there could be less of a trend or an opposite trend in Su vs PI, for a given N.
How good were the fits? [I haven't seen the actual data for either T&P or Sowers (or Stroud).] Did they control for all the relevant variables that would create apparent trends where none exist? For example, did all of their high-PI data come from a particular stratum that happened to be at a greater depth or had higher Ns? One correlation I remember, but can't find, shows the slope of Su vs N decreasing with increasing N. Was that effect involved? Were the same hammers used at all the sites, or did some get tested with doughnut hammers while others had CME auto hammers? Does sensitivity affect it, and can that sensitivity be linked with PI in such a way as to explain the trend? Were enough different sites and materials included to show that the trends are valid in general and not just between site A and site B?
Considering the mechanism by which clayey soil resists the penetration of split-barrel sampler, is there any theoretical reason why a more (or less) plastic clay with a given Su should provide greater penetration resistance?
BTW - where was Stroud and Butler 1975 published?
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Locked
- Question