Is it wise to use Geogrid and Soil-Cement? 2

[MSEdesigns]

I have a project where the owner is proposing to use a soil-cement mix in the reinforced zone of a bridge abutment. The reinforcement options are HDPE, PET or hot dip galvanized strips. I have been advised by a local distributor of the PET geogrids that the cement alkalinity will damage the PET geogrid, but what can you say about the HDPE and the soil-cement? The proposed soil is a granular type A-1-a or A-1-b and the bridge is for a 6 lane road. It has three spans of 40 ft. The reinforced walls will receive the load of the abutment, with no piles or structural footings.

 

[Boonie]

HDPE geogrids can be used with soils that have a ph range of 2-12 whereas PET is only good from 5-9. As for using soil cemented fill, you may want to investigate the interaction parameters between the reinforcement type you choose and the cemented soil. Check with the manufacturor for the applicable factors. Most of the pullout data I am familiar with deal with soil only. Good Luck.

 

[MSEdesigns]

Thanks Boonie. I think that if I use the interaction parameters of soil only when using soil-cement will be a conservative approach. What do you think?

MSEDesigns

 

[garrettk]

Just thinking out loud, but doesn't the reinforcement rely upon some movement of the soil to generate its strength? If you cement the soil, your movement will be either eliminated or at least reduced, resulting in most of the stresses being forced to the connection of the grid to the facing material.

Why do they want to do cement mixed soil behind a reinforced wall? Seems sort of redundant without knowing the details.

 

[cvg]

think of the soil cement as a low strength concrete. It won't behave like soil any more. Therefor reinforcing which is intended to be used with soil will likely not work as eluded by garrettk. you might be better off increasing the cement content, replacing the soil in the mix with better aggregate such as road base or using welded wire mesh

 

[Boonie]

The overall strength of the geogrid is based on the mobilization of the tension throughout the full length of the grid. Garrettk has a good point. If you reduce the ability of the grid to "deflect" slighlty within the soil, you will concentrate the resultant forces at the only place that will deflect i.e. the connection. I would opt to go with the tried and true (soil or crushed stone). Also, keep in mind that the basic design concept for MSE systems is the flexibility of the system. That is why we don't design MSE walls with rigid footings.

 

[Dirtguy4587]

I agree with the above - go with the soil. Just curious why soil-cement is being used if you're already planning to reinforce?

 

[MSEdesigns]

Thank you all. Very good input from you all.

Answering some questions, I believe the reasoning behind the use of soil-cement was to increase the friction angle and reduce the need of grid. Other than this I dont see any reason.

About the concentration of stress in the connection, I have some doubts. Why we talk about soil-cement as if it is not going to move at all. This is soil-cement, not lean concrete as states CVG above. The soil will move less, but it will move. Anyway, if the soil does not move as GARRETTK mentions, thus not developing the grids strength, there should not be that much load on the blocks either, since the soil-cement is not "moving". What do you think about this?

When we began producing segmental blocks in 1993 we built a 23 ft ht. wall with PET reinforcement and A-1-a and used a 3% cement to soil mix by vol and the wall has performed excellent all this years. Blocks were Keystone Standard with Mirafi grids. Reasoning behind this decision for that wall was the geotech was afraid a nearby pool could leak and cause some damage to the wall. With this he reduced the effect the small amount of water would have in the select backfill material.

Again, thank you all for your input. Let me know your thoughts on my statements above.

Thanks,
MSEDesigns

 

[cvg]

since you haven't provided properties for the soil cement, it is difficult to predict your finished strength. However, I specifiy soil cement all the time with a compressive strength of 700 psi at just 7 days. Tests on this type of product after 1 - 2 years have shown compressive strengths up to 2,500 psi. This usually requires between 4 - 6% cement. So, given the mix design, it can be quite strong and would not require a geogrid. With embankments up to 25 feet high, I have not used any reinforcement.

One reason to use the soil cement next to the pool might might have been to provide a fill that would be relatively impermeable, therefor would not become saturated if the pool leaks.

 

[TDAA]

The assumption that the grid needs to be able to move in the reinforced zone, in order to develop strength to hold the wall face in place, is wrong. Yes, in soil, the grid needs to move, to “set in” and develop the friction along its length.

But lets say it was cast into a massive block of concrete. Would it need to move? No. It would develop its strength very early on. There may be some movement right at the interface, but the strength of the material it is in would reduce the development length to get there, concentrating the load near the face of the concrete. The material would be able to resist up to its maximum capacity, and then break, just like at the connection to the block (assuming the connection str is high enough). So what would the downside be? Well, if the grid was locked in so tightly, then it is likely that the length of “loose” grid between the block and the treated soil could vary, and lead to the wall face being able to deflect differing amounts. This could result on more concentrated loads on some of the grids. As MSEdesigns said though, this is not concrete, it would be soil cement. It will still act like soil . . . an improved soil at that.

I think that is the cement is sufficient to hold the soil, the block is merely there as a facing, and the grids are shortened to just hold the facing in place, then good. If the cement does not improve the soil to that degree, and just improves the internal stability, and the grid length is determined by the global stability, then you need to look at if you are actually saving anything (i.e. can you remove some grid, go with a cheaper grid, etc.). And in the second case, is the cost of adding the cement (and all that goes with it) worth changing up the crews normal patterns during construction? Probably not. I think it gives more variables to monitor, and would slow construction, resulting in a less reliable system and more costly construction.

 

[Boonie]

Long term design strength of the geogrid is based on test data provided by the manufacturer for different types of soils. If you feel confident you are utilizing the correct reduction factors, etc. for the backfill you are using, then ok. If you don't have test data to determine the LTDS of the geogrid in the soil cement, I would proceed with caution. If the soil is so stiff that there will be no movement of the grid within the "soil", you are no longer designing an MSE structure. You are designing concrete with geogrid reinforcing. Backfill parameters and LTDS of the geogrid in that specific soil mix are required to proceed with an MSE design. If you determine these two factors,you should be ok.

 

[MSEMan]

I would guess the reason for the addition of cement has more to do with differential settlement control across the width of the abutment than with abutment stability per se.

The issue regarding mobilisation of reinforcement strength is strain related. Because the ultimate strength of a grid is only mobilised at 10% strain or more, at very low strains very little strength is mobilised. If using cement stabilised fill, you may want to adjust your "at failure" analysis for a lower soil strain (i.e. if you consider that soil failure occurs at 1.5-2% strain consider the tensile strength mobilised at that strain level.) Obviously steel reinforcement (either strips or grids) is fully mobilised at very low strain so no adjustment to the analysis is necessary (although alkalinity is more agressive towards the galvanising its less agressive towards the steel.)

Hope this helps.