Workreport 2018-15



ONKALO POSE Experiment – 3DEC Back-Analyses


Hakala, M., Valli, J., Juvani, J.



Page count:



The original objective of Posiva’s Olkiluoto Spalling Experiment (POSE) was to confirm whether or not a spalling criterion developed for granite at the Underground Research Laboratory (URL) in Canada and diorite at the Äspö Hard Rock Laboratory (HRL) in Sweden is applicable for Olkiluoto migmatite. POSE was performed in rock mechanics investigation niche ONK-TKU3 at a depth of 345 m in the ONKALO rock characterisation facility. Phases 1 and 2 were performed as a pillar experiment. Phase 1 involved the drilling of two large Æ 1.524 m experimental holes next to each other, creating an 0.87 m thick pillar between them. The pillar was heated externally during Phase 2, with hole 2 filled with diabase sand. Phase 3 was performed in a single hole at the end of the niche. After boring it was heated internally using 8 heaters and aluminium oxide filling. All three phases increased the tangential stresses on the hole boundary either mechanically or thermomechanically.

Olkiluoto migmatite exhibited shear-failure as opposed to the spalling observed in the URL granite or HRL diorite. In order to understand the observed behaviour, the POSE Phases were back-calculated using discontinuum-, fracture- and particle mechanics methods. This study made use of a 3-dimensional heterogeneous, anisotropic discontinuum 3DEC (Itasca 2017) model. The model included the geometry of the niche and holes as well as the interpreted 3D-geology with implicit or explicit foliation and a few contact surfaces. The assumed material model was so called Cohesion Softening Friction Hardening (CSFH). 

In the POSE area there are uncertainties related to the orientation of the maximum in situ compression, but the difference between the magnitudes of the horizontally aligned maximum and intermediate components is only moderate.  Based on an initial sensitivity study of the stress orientation, the in situ stress state was fixed according to existing Stress Model 1 with a σ1 trend of 112°.  This was done for practical reasons so as to be able to calibrate strength parameters as damage is a function of the stress-strength ratio.

The initial veined gneiss (VGN) and contact parameters were defined largely based on Valli & Hakala 2016 but based on true triaxial test results the confined strength of pegmatitic granite (PGR) was increased. The residual strength evolution incorporated damage-controlled (DC) tests of OL-KR10, resulting in friction being mobilised somewhat faster. The effect of deformation anisotropy with regards to damage was tested explicitly as discontinuities which had stiffness properties that were equivalent to the stiffness defined for VGN divided by an anisotropy ratio of 1.4.

Back-analysis started with thermal calibration resulting in calibrated VGN and PGR thermal parameters (increased diffusivity for both) and a heater power output of 87.5 % for Phase 2 and 75% for Phase 3. Phase 3 calibration did not require altering the thermal parameters. Reducing the theoretical heating power was most likely related to boiling and vaporization. 

Mechanical calibrations of initial strength parameter values were first done based on elastic stresses. These were then improved with elasto-plastic simulations. The limit for visible displacement was defined as strain equal to 0.5 mm of shear. Finally, the initial and residual strengths of the VGN matrix and foliation plane as well as the PGR were modified during the thermomechanical simulations so that damage formation, location and depth matched observations as well as possible. This was supplemented with a discontinuity orientation sensitivity study, to determine the effect of the angle of a contact with regards to the surface/wall of an experimental hole. 

A total of 32 cases were run, excluding initial calibrations. Cases 16 and 24 yielded the best response for the Phase 1 & 2 model, with the rock matrix strength increased by 25 %, while foliation and contact strength were decreased by 25 %. Case 16 and 24 modelled foliation implicitly and explicitly, respectively. The response to Phase 1 indicated minor plastic tensile strain in PGR, with 0.2 – 0.3 mm shear in VGN/PGR contacts. Phase 2 heating resulted in the greatest yield in the southern wall of ONK-EH2 and on the pillar sides of both holes, although lithology controlled damage could not be replicated without an unrealistic elastic behaviour of PGR. Stress paths obtained from contrasting locations revealed that at a location where damage occurred in reality, but did not occur in the simulation, simulated damage would have required optimally oriented foliation planes. The calculated strains for the strain gauges indicate that the model response is too sensitive as yield occurs too soon. The final cases for Phase 3 (30 & 32) exhibited very little to no damage at observed damage locations, although stress paths indicate that both the strength of the matrix and foliation plane are exceeded slightly at a location where damage was observed. Strain comparisons were unnecessary as the strain gauge adhesion failed in Phase 3.

Back-analysis results were a significant improvement over predictions and 3DEC proved to be able to simulate the time, location and depth of damage fairly well, although continuity is overestimated. Critical uncertainties included fixed principal stress magnitudes, the estimation of rock mass parameters based on laboratory test results as well as the post failure strength evolution, the inability of the used 3DEC approach to simulate the formation and growth of new discontinuities and finally, possible inaccuracies in the 3D geological model in the vicinity of the holes. The effect of confinement, especially with solid materials, may also have added unwanted complexity to POSE. Testing the applicability of the Hoek-Brown strength failure criterion would have provided a good comparison. Additionally, as the stress state was fixed, the calibrated strength parameters could also indicate that the maximum secondary stress magnitude may have been 25 % too low. Finally, as lithology controlled damage could not be simulated, the applied mixed continuum/discontinuum approach lacks a fundamental damage behaviour of foliated VGN related to strain localization in weaker anisotropic materials within a mixed lithology rock mass. Alternative approaches include the use of a bonded block model (BBM) or a particle flow code (PFC).

Recommendations include wire-sawing of the pillar between ONK-EH1 and ONK-EH2 to determine the true extent of damage through methods such as X-ray micro computed tomography, scanning electron microscope or energy-dispersive X-ray analysis.

Also, modification of the post failure strength evolution according to the DC unconfined compressive strength tests provided a better overall response and therefore similar tests are recommended for Olkiluoto rocks.


POSE, simulation, back-analyses, stress, damage, strain, calibration, foliation, anisotropy


WR 2018-15_web (pdf) (34.9 MB)


Share article:
This website stores cookies on your computer. These cookies are used to improve our website and provide more personalised services to you.


To make this site work properly, we sometimes place small data files called cookies on your device. Most big websites do this too.

1. What are cookies?

A cookie is a small text file that a website saves on your computer or mobile device when you visit the site. It enables the website to remember your actions and preferences (such as login, language, font size and other display preferences) over a period of time, so you don’t have to keep re-entering them whenever you come back to the site or browse from one page to another.

2. How do we use cookies?

A number of our pages use cookies to remember your actions and preferences (such as login, language, font size and other display preferences.)

Also, some videos embedded in our pages use a cookie to anonymously gather statistics on how you got there and what videos you visited.

Enabling these cookies is not strictly necessary for the website to work but it will provide you with a better browsing experience. You can delete or block these cookies, but if you do that some features of this site may not work as intended.

The cookie-related information is not used to identify you personally and the pattern data is fully under our control. These cookies are not used for any purpose other than those described here.

3. How to control cookies

You can control and/or delete cookies as you wish – for details, see You can delete all cookies that are already on your computer and you can set most browsers to prevent them from being placed. If you do this, however, you may have to manually adjust some preferences every time you visit a site and some services and functionalities may not work.