Micropile test - What length contributes to capacity? 2

[Settingsun]

Hi all,

A micropile contractor has proposed a test to determine the skin friction provided by a particular ground layer. They have installed a micropile as follows:

- 219mm (8.6") steel tube casing used to drill to 25m (80') depth.
- 40mm (1.6") threaded steel rod down the centre of the tube. All except the bottom 3m (10') is sheathed to debond from the grout.
- Grouted to top of casing with "good flow to displace soils".
- Casing withdrawn with grout topped up along the way except for very near the end of casing withdrawal.
- Grout to achieve 50MPa (7ksi) before test.
- Tension test to be done with the jack placed on ground directly above the pile (on steel plates). I understand it isn't bearing on the pile as the grout didn't go to surface.

There was no borehole at the test location but it is expected that the pile is in 15-18m (50'-60') of soft clay and loose sands, overlying 7-10m (23'-33') of medium-dense sand.

Because only the bottom 3m of the threaded rod is bonded to the grout, the contractor is saying that the test load achieved can be attributed entirely to the bottom 3m of medium-dense sand. This doesn't seem right to me. Since the grout installation is essentially the same over the full length, the top 22m should be equally bonded to the ground as the bottom 3m. Since the load will be applied to the base of the micropile by the steel rod, the grout column will be in compression and capable of transmitting load to the upper ground layers (in my current opinion).

Appreciate comments on this. If only the bottom 3m is contributing to geotechnical capacity, what is the reason/mechanism that prevents the upper layers contributing?

I understand that counting only a part of the micropile length (in the target resisting ground layer) is a common and conservative *design assumption* but from what I've read it's not a reflection of the actual action of a micropile.

Thanks in advance.

 

[oldestguy]

Sounds goofy.
You need a geotech involved. Normally a recognized pile load test needs an actual test boring nearby a proposed pile installation. That installation can be tested with a standard load test procedure as specified by ASTM or other recognized engineering group. Finally any load test usually will then be used for design with a safety factor to allow for such things as the soft clay settling later,etc. You might even use two load tests, one full depth and one only in the soft clay. However ahead of all this a conventional geotechnical investigation and report will have lab tests on the materials and use the testing done on the soils during the geotech investigation. I'd do that first before any load test is considered. During pile installation there also is the ability to get "test" information usable to direct how deep to drive, etc. You may not need the load test.

Edit: It's always possible that later on the clay zone applies a negative load. The geotech can advise on this aspect. You need full soil properties.

 

[Settingsun]

Thanks, Oldestguy. I've also found a few references that say the test pile will act as I thought.

 

[MTNClimber]

Just because they debonded the bar from the grout in the upper layers does not mean that grout in the debonded zone isn't going to add resistance. It can actually add a lot of resistance. They were better off leaving the casing withdrawn 3m from the pile tip since the steel to soil bond won't be as much high.

I've seen engineers spec that the contractor needs to washout the grout above the bond zone, but I'm not sure how accurate that is since you could over or under wash out the grout. I prefer to put strain gauges along the pile to get an idea of how the load is being transferred along the length of the pile.

 

[Bobby46]

There's several failure modes for something like this. One failure mode is the grout/ground interface. This is usually assumed to be the bonded length of the anchor. It is verified by load tests that approximate the unbonded length through deflection measurements. Another failure mode assumes that a wedge of soil is ripped out of the ground. This considers the entire length of the anchor.

Some references with good information are:
The Post Tensioning Institute Recommendations for Prestressed Rock and Soil Anchors
FHWA Geotechnical Engineering Circular 4 Ground Anchors and Anchored Systems
I think there is a USACE manual on anchors for dams but I don't remember what it's called.

A side note: I think 3m is pretty short for a bonded length. Maybe your local soils allow for a short length.

 

[MTNClimber]

Bobby46, I think they're trying to test the skin friction resistance (grout to ground) of one specific soil layer, the medium dense sand layer. This is usually done on jobs where a more accurate pile design can save the owner money by reducing pile lengths.

 

[Bobby46]

But wouldn't the load test show this? PTI sets limits for the Apparent Free Stressing Length measured during load tests. If the load test indicates that the Apparent Free Stressing Length is between 80% of the Unbonded Length assumed in the design and 100% of the Unbonded Length plus 50% of the Bonded Length assumed in the design, then the anchor is acceptable. Isn't this verifying that your anchor is achieving it's capacity at the depth/soil layer you design for?

Alternatively you could place a tell tale at the top and bottom of the 3m layer of sand and use PL/AE to solve for the load resisted in that soil layer

 

[MTNClimber]

I'm not sure I understood the OP's concern but what I gathered was that they debonded the center reinforcement bar from the grout except for the bottom 3m. Their theory was that the pile will only use the bottom 3m of grout to resist the load during the load test. The way I see it, the pile will still use resistance from the unbonded zone since the grout from the bonded zone is still connected (not as well) to the grout in the undbonded zone. If that is the case, then it leave everyone thinking that the grout-to-ground resistance of the 3m bonded zone is higher than it actually is since it will take more effort to fail the pile than if it was only truly grouted in the 3m zone.

I agree, tell tales would of been nice to have.

 

[fattdad]

apply the load to the threaded bar only.

then again, maybe I'm confused?

f-d

ípapß gordo ainÆt no madre flaca!

 

[MTNClimber]

Think of the unbonded zone as a grout plug. It's going to have resistance even when you're pulling from the bonded zone.

 

[Settingsun]

Hi all,
MTNClimber is thinking the same way I am.

As far as free length goes, it will measure quite high since I have no doubt the central steel bar will stretch over 22m of its length as it is debonded. The load is applied to the steel bar (only) at the pile head. The load isn't transferred to the grout until near the toe of the pile, by steel-to-grout bond.

The question is as MTNClimber said: Once the load is in the grout column (near the toe), will the grout be restrained by only 3m of soil or by the full 25m? Or put another way, do you divide the applied load by 3m or by 25m to calculate the average bond between grout and soil? Huge difference obviously.


---
EDIT: I've found in the 1989 version of BS 8081 "Code of Practice for Ground Anchorages" the following statement (section 11.2.3.1):

"Where a single phase grouting operation for both fixed and free length is planned, the two grouted sections should be separated by the attachment of a packer or other compressible medium around the tendon at the top of the fixed anchor length. The packer should prevent transfer of load from the fixed anchor into the free length grout during stressing."

This suggests that you do need to guard against the supposed free length affecting the test result.

 

[BridgeSmith]

Seems to me the only way to measure the geotechnical (ground to grout) resistance of just the 3m bonded zone is to only grout that zone. Otherwise the grout in the bonded zone is restrained by the remainder of the grout column, and you're testing the geotechnical resistance of the entire 25m or the structural (bar to grout) resistance of the 3m section, whichever is less.

 

[MTNClimber]

A packer would be a good idea. I never understood why literature I reviewed didn't really address this. I guess it's more common to just do a full scale load test instead of checking the geotechnical resistance of a specific layer. But those presumptive values in PTI had to come from somewhere.