Finding the ‘G’ spot – High Pressure Dilatometer
I’ve been recently managing ground investigation works at a car park that will become a temporary construction site for four large steel frame modules (Approx. 50x50m). The GI is for the temporary foundations which will support these modules. The cores that have been found on site so far have revealed Weald Clay Mudstone at the depth of the proposed foundations. Recently I observed the use of a pre-bored High Pressure Dilatometer (HPD) which I had not heard of before. I thought this might be useful to explain a bit further. It’s an unusual but accurate method of finding horizontal soil strain ‘ε’ and shear modulus (G) in-situ instead of laboratory triaxial tests. It’s used much more commonly in France and other parts of europe but I’ve been told it is increasingly being used in the UK. It was developed from the Ménard pressuremeter, which is still used but is less accurate. It can only really be used in fine grain soils and most rocks, though problems are encountered in limestone where there’s lots of flint. The device isn’t useful in coarse grain material due to the low plasticity of the soil giving inaccurate results. It works by placing a cylinder at a certain depth with a membrane surround (see image below). The membrane inflates to apply pressure against the soil, where a number of sensor probes measure the pressure of the soil. This pressure is plotted onto a graph against the length of membrane expansion.
It comes in 3 different types:
- Self-Boring – Has the advantage of being more accurate as the dilatometer fits the hole much better and there is less disturbance of the soil, providing more accurate results. Not viable where there are layers of coarse grained material that can disturb the borehole or where rock is encountered.
- Push type – Similar to self-bored, but the sensor is driven into the ground. Has similar advantages but is limited to soft fine grained material such as clays. This can raise the stress of the surrounding soil, limiting the data you can get from this method. It is however much quicker.
- Pre-bored – A borehole is made to the appropriate depth and a dilatometer inserted into the pre-formed hole. The main advantage of this is that it is more versatile and can be used in a number of different type of soils. Best used in rock and stiff clays. Cores can also be taken during boring enabling lab testing of samples. Less accurate due to soil disturbance during boring.
A graph of average readings from the car park soil can be seen below showing cavity pressure against cavity strain. You will notice that there are a number of loops on the graph not too dissimilar to the stress/strain curve. The HPD operator expands the membrane until the pressure begins to level, and then releases some of the pressure, allowing for the soil to recover. The pressure is then reapplied causing an ‘elastic rebound’. This is usually done three times to give a spread of results. As I understand it is the average rate of this elastic rebound that can be interpreted to find ‘G’. The test is continued up to a point where the pressure begins to level off. It is not so easy to see in this image, but the individual sensor results show a gradual levelling off. I was told by the operator that they usually stop the test at this point as there is a risk of membrane rupture. As each of these cost about £80k this is understandable!
I have been told there are relatively few GI specialists in the UK who can carry out this kind of test and even fewer who are certified to interpret the results. That said it seems this is a much more accurate way of finding undrained soil properties from in-situ testing against more empirical methods such as soil classification properties against previous data. It may be particularly useful on sites where there is little empirical evidence. It is also more accurate than triaxial testing as the soil is not disturbed significantly. The HPD is likely useful for projects where a high degree of accuracy is required from the GI to build substantial foundations reducing the risk associated with the ground. This said I think that the HPD might not have been required on the this site as the foundations are temporary: a less costly triaxial test may have sufficed. I would be interested to know what other opinions might be.
Other information on soil properties can also be derived from these results such as the coefficient of uniformity (Cu), though for brevity this is not included above. Further reading can be found in Craig’s Soil Mechanics 8th Ed p240-8
Holdfast Training Services Trial Blog 
Al, you mention that this equipment is best suited to fine grained soils and rock. Do you know if the GI subcontractor already have access to site borehole data confirming the soil type (historic or a two staged GI) or did they mobilise the equipment based on assessments from geological mapping and ‘experience’ of the types of soils encountered in SE England?
Yes the site had been used before as a temporary site to build a large airbridge in 2005/6 so there was some existing GI data from that project. Much of Gatwick sits on a Weald Clay/mudstone layer at the depth of the proposed foundations. This has been shown by numerous bolehole reports going back several decades. Therefore it would be reasonable to expect to find this kind of soil. I also forgot to mention that a CPT had been carried out at the borehole sites the week before which also gives a good indication of the type of soil (though only down to about 5m when the clay became too stiff for further penetration).
Yep a nice piece … we don’t cover dilatometer in the RE course for fairly obvious reasons
I’ve never used a self boring or a cylindrical one just the older style Menard ( plate) apparatus
Your kit expands laterally and measures the volumetric strain. In the elastic region this measures the Bulk modulus and then elastic theory is used ( with a bit of assumption on Poisson’s ration to get shear modulus(G) The reason why this is of interest is tat drained and undrained shear modulus are identical whereas the related Young’;s modulus differs for fine grained soils
The point is that this is high end measurement of soil stiffness and then question is why?
It must be to assess likely settlements using Finite Element modelling ( I would think)
How you get the coefficient of uniformity from Pressurmeter tests is beyond me….!
John, my understanding is that the high end measurement was added into the scope of the GI in order to mitigate any risk in the ground as much as possible. The designer would be to be able to design more efficient foundations and determine soil settlement. Given that there are to be four sets of large pile foundations costs could be saved on the foundation design. Furthermore, the CPT was only able to penetrate to around 5m, whereas the HPD can find soil properties at a greater depth. As the foundations are required to take high loads from the steel structure of each module, each pile will likely need to be installed at a depth greater than 5m.
From what I’ve read Cu can be found by plotting cavity pressure against the logarithm of volumetric strain in the cavity (ln(dV/V)) and then using the gradient of the curve of this plot will give Cu. I will not pretend to know how this is derived as it is beyond my understanding. It is detailed in Craigs soil mechanics 8th ed p 245-6.