Underground Excavation Support Design - In-Situ Horizontal Stress

[palmahouse]

I am designing support systems for underground excavations, and have been using a stepped evaluation approach presented in Underground Excavations in Rock (Hoek and Brown). The procedure is a good guideline for my projects, however, I deal with underground excavations that are typically less than 70 feet deep (maximum) in tectonically altered volcanic rock on the west coast of the U.S., and Hoek and Brown's focus is on deep mines.

The technical hurdle I encounter is estimating the in-situ horizontal stresses in the rock. This factor has a significant effect on the support system design (independent of the design/evaluation procedure).

I am tempted to use a k-value of 1.0 to estimate the horizontal stresses for my projects, because some literature I reviewed indicate that k = 1 is probably a good estimate for sheared rock.

Please provide comments regarding estimating the k-value for sheared volcanic rock (typically tuff, andesite, and rhyolite) within 100 feet of the ground surface.

 

[ishvaaag]

I have looking for answer to your question and found the likes of Hoek approach that you mention, say...

http://www.rocscience.com/hoek/pdf/10_In_situ_and_induced_stresses.pdf

Also found that awesome residual stresses can be locked even at the surface itself; but this are locally relieved upon excavation; this has made me realize that you have a vertical soil support problem. For these problems

Deep Excavations, a practical manual, 2nd edition
Malcolm Puller
Thomas Telford Limited 2003
p.149

states a frame to treat weal rock with a phi' as follows

Chalk 35º
Clayey marl 28º
Sandy marl 33º
Weak sandstone 42º
Weak siltstone 35º
Weak mudstone 28º

then you can if wanting to apply the at rest value deduce sin phi' from that 1.

K0=1-sin(phi')

and lesser venturing for active situations.

Other thing is that stress relief apart, a hardback of rock is something upon which stresses can build; so if you would let say water infill between your wall and the rock it might develop fantastic pressures.

 

[edmy]

The K-value depends on the geologic formation of the site. You could have k>1. Please refer to the book "Introduction to Rock Mechanics" by Richard E Goodman (1989)

 

[BigH]

Besides the horizontal stresses - do you have joint planes that are of concern? support needed to prevent "pop outs" of unfavourable jointing.

 

[palmahouse]

Thank you all for your insightful responses.

jshvaag: I think that assuming a phi angle and calculating k-not leads to underestimating horizontal stresses in rock in some cases (but maybe not in all). Sure, excavation relieves the stresses, but, the in-situ horizontal stress still effect tunnel liner design, and the methodology I am using does incorporate rock stress relief caused by excavation.

Good point on the potential for water pressure. I am not so sure that the procedure I am using would properly incorporate hydrostatic pressures imposed on the liners (the procecure may only lower the shear strength of the rock mass from water pressure in the joints). For this project, we have a relatively dry hill.

edmy: Does Goodman's book provide a good discussion on in-situ horizontal stresses in young rock, as well as measured data? If so, that may be my best bet because he mostly studied the geoolgy I am dealing with (in Northern California). I was his student, so, perhaps I will call him (if I can track him down).

BigH: In this project, wedge failure potential is not a concern. The joints are all near-horizontal. We had only one minor roof fallout during excavation that exposed good and stable rock. I like to stabilize wedges with bolts instead of increasing stiffness of the liner, because I think it is more efficient and reliable that way.

 

[priyainuk2009]

I did some works for a drainage tunnel in Hong Kong.The tunnel runs through predominantly Tuff rock mass. Hydrofracture surveys confirmed that average k varies from 4.0 at 40m depth to 2.0 at 120m depth.