Load Transfer to Foundations of Buried structures

[VAD]

Much needed. Thanks for the invitation. I will start by posing a question that has implications on work that geotechnical engineers are involved with with but never seemed to have been well explained.

Do we truly understand the phenomenon of arching in buried structures and how it influences the loads transmitted to structures such as large culverts?. There is very little said about soil load transfer to buried structures without base slabs that rest on footings.

If one examines the AASHTO and Canadian codes one will note differences in the approach in determining the loads. Although many apply the forulae presented in these codes there is very little available to discuss the mechanisms of load transfer. in fact some use the loads providede by the manufacturers of these structures who seem to provide the structural aspect but leave the responsibility of the foundations to the geotechnical engineer. All concepts lead to Terzaghi's work on arching which is in his Theoretical Soil Mechanics text where he references work of his predecessors. Work by Martson and Spangler which can be understood is now often put aside by statements such as that this approach is not applicable today. No reason given in particular.

A particular case in mind is how do we treat a footing projecting outside of the line of the barrel of a multi-plate culvert structure. Do we account for the overburden pressure in the design of the footing as an additional load on the on the projected part as height x density or do we reduce this amount as a result of arching or do we neglect it completely. The AASHTO codes indicates this should be accounted for but no mention by the Canadian code. In fact in the Canadian code there is very little information regarding the design of footings below these structures. Comments, opinions would be welcome. Since we apparently are governed by Codes we should strive to ensure that what is said is clear and understandable.

 

[jheidt2543]

VAD:

A very timely question. I just downloaded the paper "Design Loading on Deeply Buried Box Culverts" by
Kyungsik Kim1 and Chai H. Yoo, Journal of Geotechnical and Geoenvironmental Engineering, Vol. 131, No. 1, January 1, 2005. It is available at ©ASCE, ISSN 1090-0241/2005/1-20–27/$25.00.

They cite two papers by Martsen and two by Spangler, but none by Terzaghi. There is a pretty extensive list of references too. They discuss many of the points you raise so, that paper may be of some help. Here is the abstract directly from the paper:

"The current American Association of State Highway and Transportation Officials (AASHTO) standard specifications for highway bridges and AASHTO LRFD bridge design specifications stipulate the computation of the design load on the box culvert primarily based on research by Marston and Spangler. Although this procedure may be applied conservatively for most ordinary culverts, an opportunity exists to evaluate a more realistic soil-structure interaction behavior based on modern finite element analyses of deeply buried concrete box culverts. The Duncan soil model, represented by hyperbolic stress–strain curves, has been used for properties of backfill and in situ soil. The backfill heights are varied from 15.2 to 61.0 m s50–200 ftd for the embankment condition and 15.2–45.7 m (50–150 ft) for the trench condition. An optimum combination of parameters has been identified for use in the imperfect trench installation method. The data from several hundred hypothetical models with various parameters under three typical installation methods, i.e.,embankment, trench, and imperfect trench installation are characterized and quantified using regression analysis."

Sorry of the long post, but I thought quoting the abstract would give you the best summary of the paper.

 

[VAD]

Thanks jheidt2543:

I stopped my subscription this year to the ASCE journals as I can access these at the local university library 5 mins away from home. Will check that paper out. Terzaghi's paper dealt with arching of sand over a trapdoor.

Thanks again

Regards

 

[jheidt2543]

Perhaps Saddam Hussain read that paper?

 

[BigH]

An interesting question - and a hard one for the "first" thread! Not really thought about it much as most of culverts/pipes I have are relatively shallow. This would be a good thought for some of us to do a review. I typically see that methodology for pipe culverts below embankment is to build the embankment up some 2 metres above the pipe culvert, then dig out the trench, install the pipe, backfill and compact about the haunch and for 1/2 pipe dia above the pipe - then loose fill in zone up to embankment - then continue with embankment. This approach to using non-compacted fill has always struck me as funny - in China, they would just dump in the fill to the trench all the way to the road, then build add a bit more fill above until traffic and self-settlement caused it to settle - then top up as needed. I'll have to give some more thought - VAD - can you cite the various references you've quoted; over here I have hard time with getting such - I've an AAHSTO bridge code 1996 with supplements through 1998. The Canadian code - is it the Ontario bridge code?

 

[VAD]

Hello Big H.

The references are AASHTO LRFD Bridge Specifications 1998 and Canadian Highway Bridge Design Code CAN/CSA-S6-00.

At the moment I am working on some aspects which I believe would make it much clearer to the application of the relationships presented and foremost illustrate how the geotechnical aspects should be addressed. This is one area where very little guidance has been presented regarding the geotechnical input yet so much depends on that very aspect in the design assumptions. It will take some time but seems promising so far.

Another topic is negative skin friction, drag load, downdrag etc. Although much have been said on this topic there are still areas that are not well explained and we all use relationships slavishly. That will be another topic later.

Jheidt 2543:

I acquired the reference and it took me about 2 minutes to realize that the paper does not address the needs of the practicing engineer in relation to fundamental understanding of load transfer although there are seemingly good results achieved. One statement made that all buried structures in the embankment condient are yielding structures told me that resaerch many times forget that the situations in the field. The statement is a way that some researchers get around avoiding the complications in the real world. Nothing wrong with that but what is needed is clearly a reasoning so that the reader can utilize the findings of the research intelligently when he/she is faced with a problem to solve. Very often it is also lack of the field experience. These are my personal opinions.

Regards

 

[BigH]

VAD - another aspect to consider is the building of a design camber to offset the foundation settlement; not precisely a load transfer point, but amazing how many engineers I talk too don't even think of this - and it is somewhat tricky if you are adding, say a 15m embankment to an existing one and extending its culvert on soft soils! Thanks for the code sections - I'll write Toronto and see if they can send them to me.