Wondering if anyone has any tips or advice on how they approach calibrating hybrid bolt structural elements.
I’m aware theres an online tool available from Itasca for calibrating the bolt properties (Calibration Tool). From experience the bolt behaviour is largely driven by the discretisation of the bolt element (segment length) relative to the size of the zones in the model which the online calibration tool cannot calibrate for as this is model specific. In the case of DFN models which contain rock joints, the dowel length also becomes critical as it needs to be chosen such that the bolt element is discretised with sufficient structural nodes on either side of the rock joints that are intersected by the bolts.
Is it simply a case of running the model with various bolt discretisations and verifying the adopted segment/dowel lengths are appropriate?
do you have shear/pull test data that you’re able to calibrate to?
have you seen the manual section here regarding the effect of node spacing and the effect on stiffness and hence shear response of the bolt?
my understanding is that you need to select a node spacing small enough that it does not affect the overall bolt behaviour (i.e. as you say node spacing is small enough to be relevant to your zone size and joint spacing). however, be aware that decreasing the node spacing will increase the overall bolt stiffness and whilst you can adjust the input shear and axial stiffness properties of the bolt to overcome this, it can’t be “calibrated”.
i’d suggest:
calibrating a simple model to shear/pull test data with your best estimate at required node spacing based on what you know your zone size and joint spacing will need to be in your final model.
then using calibrated parameters, run scenarios of your actual model varying only different bolt node spacings, to validate behaviour.
you may then need to iterate (1) and (2) above, revisiting initial calibration with a different node spacing, noting above comment on reduced node spacing increasing stiffness, which you might want to account for.
Thanks for your reply. I agree, it is best to run a simple model to calibrate the bolt parameters before applying it to the full model.
On the joint spacing, this is something I’ve always had trouble with estimating as it can vary greatly throughout the DFN domain. I was thinking to extract small domains from the full DFN model to run as simple models for calibrating the bolt discretisation. You would obviously have to sample different domains from the full DFN and run multiple tests though to capture the range of joint spacings and bolt discretisations. Is this something you have tried in the past?
assuming your excavation is a 3D geometry and your bolts are perpendicular to the excavation face, i don’t think you’ll ever calibrate this as your bolt orientations will vary with respect to the joint orientations and therefore even in the simplest case, “apparent” joint spacings will still vary. however, based on your DFN fracture intensity, you should be able to estimate your required node spacing/segment length? note there is a maximum-length argument to the structure hybrid create command, not just the segments argument: here.