By Louis Marcil
This is Part 3 of a series of posts about pressuremeters. Parts 1 and 2 are available on our website here and here.
There are two main approaches for using PMT results: the indirect and direct methods.
The indirect methods consist in estimating fundamental soil parameters from PMT test results, and using them, along with conventional design methods, to better understand and predict soil behavior. These methods rely on theoretical bases by referring to different mathematical representations of the behavior of the soil under load.
Direct methods rely on the similarity between the behavior of the soil with respect to a foundation and during a PMT test to claim that one can use parameters specific to the PMT test (pressuremeter modulus and limit pressure) to directly predict the behavior of a foundation. These methods have theoretical bases, but are also empirically adjusted. They are mainly associated with PBP testing using hydraulic probes.
Both methods come with strengths and weaknesses, and have applications under different circumstances.
The most frequently obtained parameters with the indirect methods are the soil modulus, lateral stresses and undrained shear strength, and in some cases the pore water pressure, effective cohesion and angle of friction.
The soil modulus estimated from a PMT is a particularly valuable parameter. It can be used in elastic or finite element analyses to estimate settlements and lateral deformations. The main challenge to this method is determining which modulus to use as an input, considering that the soil modulus is affected by several elements (strain level and rate, stress level and history, water content…). Reload moduli are often used in order to mitigate the effect of soil disturbance. A field for which pressuremeters (and flexible dilatometers) are very useful is the design of large structures sitting on rock where pressuremeter moduli can be used for estimating rock mass deformations through finite element analyses.
Provided that the drilling and insertion of the probe are done properly, the pressuremeter test is one of the best tools for estimating in-situ lateral stresses. Different methods have been suggested to derive this parameter directly from the PMT curve, or for estimating it from empirical correlations. Different methods were also developed for obtaining an undrained shear strength from PMT test data. One of these consists of inverse modelling analysis whereby analytical models are compared to the PMT data in an attempt to determine strength parameters.
Often these indirect methods will be used as complementary tools during large / high risk projects where more resources are justified. PMT tests will be all the more justified in cases where soil parameters are difficult to obtain by conventional means, for instance in weathered or fractured rock or weakly cemented material.
Some parameters and the methods suggested for obtaining them, are sometimes directly associated to a specific type of pressuremeter (e.g. electronic probes and SBP). In general, moduli of deformation, ultimate pressures, undrained shear strengths and, to a lesser extend, lateral stresses can be obtained from all types of pressuremeters – but often using different methods.
While indirect methods can be extremely useful, they come with some weaknesses that need to be considered. First, as mentioned above, these methods have often been developed from the results of different types of pressuremeters and test methodologies, which are not standardized. Any variation in the type of equipment and methodology may affect the final results. Also, some of these methods may be pretty complex. They use different soil constitutive models depending on the type of soil and estimated parameters (linear elastic perfectly plastic material with or with no volume change, parabolic or hyperbolic PMT test curve….), and are sometimes based on assumptions which are difficult to verify. Soil behavior is complex, variable, and difficult to model in a perfectly representative way. Furthermore, the effects of the testing operations (loading sequence), soil disturbance (also present with SBP) and heterogeneity can exceed the theoretical benefits of using a more complex vs. a simpler model. Efforts are made to compensate for these effects, but it is not clear the level to which these efforts are successful. Finally, on a more elementary level, these methods are often based on the results from instruments (electronic SBP) that are not as commonly used because they are less versatile (in terms of testable soils), more difficult to handle, and commercially less accessible.
Direct (or semi-empirical) methods are often used with more confidence for the design of civil engineering works. They can be considered for several applications. Settlement and ultimate bearing capacity can often be reliably predicted using relationships initially developed by Menard and involving pressuremeter modulus and limit pressure. These direct methods, initially developed for shallow foundations, are being extended for deep foundations thanks to recent developments that better account for skin friction. But the more predominant the shear deformation over skin friction, consolidation, and other elements, the more the settlement predictions based on PMT will prove representative. Also, it is interesting to note how more representative the bearing capacity estimated from PMT is as compared to the one deducted from conventional methods involving the cohesion and the angle of friction. The use of the PMT has allowed in some places (Chicago, Toronto for example) to significantly increase allowable bearing capacities.
PMT-derived methods for the design of laterally loaded structures prove to be very representative, which is not surprising considering the close analogy between the behavior of these structures vs. the loading pattern of a PMT probe. PMT can also be used for pavement design and soil improvement control.
There are several reasons for the effectiveness of direct methods. First, these methods are less complex, less theoretical, and more grounded to ‘real life’ geotechnical behaviors of foundations. They have been adjusted and validated by a larger number of case studies over decades. They are part of a standardized ‘complete system’ that specifies which type of pressuremeter to use, how to run the test, and how to use the results. Furthermore, they are based on results from test equipment that is versatile (PBP can be used in almost all types of soils and rocks), easier to use, and commercially more accessible.
But as is always the case, these methods also come with drawbacks. First, their simplicity and their ‘autonomous’ aspect, which, on the one hand, are an advantage – a user can quickly predict a settlement and bearing capacity from a simple calculation based on the PMT parameters – makes them in a way partitioned. Thus, they do not use fundamental soils parameters and associated design methods, and remain relatively unknown. Also, they require empirical adjustments for the soil and foundation types, and for the inaccuracy of the theoretical model. This will be disheartening for some, especially when testing for less common soil types that might be more difficult to associate with the categories defined in semi-empirical methods. Finally, the level of remolding, inherent to drilling and probe insertion, must be controlled using specific drilling methods and validation tools.
Overall, as seen above, PMT tests can provide valuable information for a large number of applications especially for shallow foundations, laterally-loaded foundations, high-rise buildings, large structures on rock, and for soil improvement control. It is important, however, to note that two approaches have developed in parallel, which has contributed to an increase in the number of device types, test procedures, and interpretation methods, and has led to some confusion. This is especially the case when designers are presented with results from equipment and techniques that mix the two approaches. Ideally, the aim should be to achieve a higher degree of standardization of equipment, testing and interpretation methods, and reference databases. That will require time and the involvement of several stakeholders. In the meantime, communication and training should be emphasized in order to minimize this confusion and promote the proper use of this type of testing, which can provide high quality information for soil and rock behavior.g-am