Tremie concrete to steel pile bonding strength

[Guest090822]

thread255-380420

In designing a concrete tremie slab for a cofferdam I am coming up with a really thick slab. One of the geotechnical engineers I work with said that I can reduce the thickness by taking advantage of the bond between the tremie concrete and the new piles. He also indicated that some engineers account for the bond acting on the perimeter of the sheet pile wall as well.

I'm comfortable taking advantage of the tremie concrete bonding to the new piles, but have concerns that using the bond of the tremie concrete where it meets the sheet pile since I don't think the sheets are "clean" and a good bond may not develop.

The state of Indiana has guidance for determining the resistance of the bond with the piles. They base it off of a 36 psi bonding strength and if the depth of the tremie is greater than the pile depth "d" (From AISC manual - not the embedment) then only allow you to use "d" times the perimeter of the pile times the 36 psi bond strength. I just don't think it is a good idea to apply the 36 psi bond to the perimeter of the cofferdam sheets as I think that is somewhat unreliable, but at the same time if it seals out the water then a bond must be developed. Maybe use 25% of the strength or something like that?

I should note that the soil has some boulders and there is a dense layer of material not too deep. The boring logs indicate that the H-piles can be driven, but the sheets can't be driven deep enough to prevent seepage/piping.

Any suggestions?

 

[SlideRuleEra]

Under the best conditions Bethlehem Steel recommends a bond strength of no more than 20 PSI between tremie concrete and steel H piles. Under less favorable conditions no bond strength (0 PSI) is recommended. For the details of their recommendation see page 36 of "Bethlehem Steel H Piles", located on this page of my website: SlideRuleEra.net

You mention that the calcs indicate a "really thick slab". Of course I don't know what thickness that refers to, but a seal that is 10 or 15 feet thick (or more) is not uncommon. IMHO, all by itself, the seal should be thick enough (i.e. heavy enough) to prevent theoretical cofferdam flotation. Any bond strength between the tremie concrete and H pile may be used as part of the flotation safety factor... but not essential to hold the cofferdam down.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[PEinc]

I often see contract plans where the consultant shows a "tremie seal" that is only about 2 feet thick. IMHO, this is usually too thin to resist hydrostatic uplift pressure. I agree with SRE that real tremie seals are usually thick, easily well over 10 feet thick. Don't forget a safety factor when determining the thickness of the tremie seal concrete. I also would be concerned about the bond of the tremie concrete to the SSP which most likely is not clean steel. I doubt anyone is sending down a diver to clean the SSP before placing the tremie seal.

http://www.PeirceEngineering.com

 

[Guest090822]

The project in question consists of a 48'x16' in plan sheeting system with a head of 12'.
We use 145 pcf for tremie concrete (140 more common). Considering submerged concrete the relationship can easily be made that the thickness of the tremie seal is 0.76 x Head (no factor of safety).

I came up with a (0.76)(12')= 9.12' x 1.5 FS = 13.7 ft.

My geotech guys took into account the bonding of the 21 - HP10x57 piles and reduced it to 4 feet thick. His anaysis said he only needed 1.7 feet but he bumped it
to 4 feet.

I've seen the bonding strength range from 20-36 psi depending on the source.

I'm more concerned with what is standard practice in regards to justifying a reduction in the slab thickness. 14' vs 4' is quite a difference but they seemed to indicate everything will be fine.

I'm a bridge designer and have limited experience with cofferdams. I'm just trying to learn something new.

Thanks for the replies!

 

[appot]

Maybe I am missing something obvious, but where is your 0.76 coming from?

12 ft of water head * 62.4 = 750 lb (I typically ignore the per ft^2 term by assuming a one square foot area)

750 lb / 145 pcf concrete = 5.17 ft of concrete

5.17 * 1.5 = 7.76 ft


Edited to include 1.5 FS

 

[Guest090822]

145 pcf - 62.4 pcf= 82.6 pcf submerged density. 62.4/82.6 = 0.755 rounded to 0.76

Since the tremie concrete has the higher density the relationship is for every 1 foot of water head you need .76 feet of tremie concrete if placed in a wet condition.

Factor of safety added afterwards depending on temporary or permanent condition. In the state of NY we typically use 1.25 for temp and 1.5 for permanent.

 

[SlideRuleEra]

TheRick109 - For a 4' thick seal, I went through calcs to see how much uplift each HP has to develop. Like PEInc, I ignore bonding to the sheeting.

In summary, assuming all 21 HP are uniformly distributed, each HP has a tributary area of 36.6 ft².
Each HP has to withstand 15.3 kips of uplift. That is the actual force, no safety factor. Per both the Indiana formula and Bethlehem Steel, the bond strength will do it. Wonder if the HP friction with the soil is up to the job?
My calcs are attached.

IMHO, a better plan is a 10' thick seal then taking credit for the HP / Tremie Concrete bond as the safety factor.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[PEinc]

Unreinforced, tremie concrete probably weighs about 140 pcf or 140 - 62.4 = 78 pcf submerged.
12' head (from water level outside cofferdam to top of tremie pour) = 12' x 62.4 pcf = 750 psf.
750 psf / 78 pcf = 9.6' of buoyant concrete.
9.6' x 1.5 FS = 14.5' of tremie thickness. Use 15' (or less if you add in the bearing pile contribution).

http://www.PeirceEngineering.com

 

[Guest090822]

The 12' of head is to the bottom of the proposed footing. Do they normally pump the water level down to that level before placing the tremie slab, or is it typically done with a lower or mor head differential? The sequence of operation is probably something I'm not taking into consideration.

I do appreciate the replies - you both have been very helpful.

 

[SlideRuleEra]

The seal has to be placed underwater (using a tremie), before the interior of the sheeting is pumped. The seal has several purposes:

1. It "seals" (i.e. waterproofs) the bottom of the cofferdam to keep water from entering the interior when it is pumped.

2. Weight of the seal keeps the dewatered cofferdam from trying to float... and if it did try to float, it would collapse catastrophically, likely with fatalities.

3. The seal resists horizontal pressure (at the bottom of the cofferdam) on the sheeting when the cofferdam is dewatered.

I see you do need information on the construction techniques. You are in the right place, PEInc is one of the top practicing authorities in the USA. I comment more from a different view point.

Read the attached (old) documents for a basic overview, then come back here for discussion.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[Guest090822]

SlideRuleEra - I read through the documents, they were great.

I also have a book I found titled "Construction Dewatering - New methods and applicaions, 2nd Edition" by J. Patrick Powers (1992) that gives the thickness of the tremie using the submerged density though and would basically accomplish the design as PEInc suggests.

If dewatering to the bottom of the tremie concrete, then I can take advantage of using 140 pcf which will have the largest impact on the thickness required. That is something I hadn't considered.

Taking that approach and using the piles I get a factor of safety of 1.5 or greater depending on the pile method so it should be fine.

Thanks again.

 

[SlideRuleEra]

If dewatering to the bottom of the tremie concrete, then I can take advantage of using 140 pcf which will have the largest impact on the thickness required.

That's not how it works.
What thickness seal did you come up with?

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[SlideRuleEra]

TheRick109 - Sorry, I did not intend to confuse the situation. I don't use "submerged weight" for calcs, but account for it by using Archimedes' Principle.

The key fact is the seal concrete is below the water table and is buoyed by it. There is no way around this.

Archimedes Principle is easier to illustrate than to put into words. Here is the diagram posted in response to another member's question on this same subject:

No matter which way you perform the calcs, submerged weight or Archimedes's Principle, for the 12' depth a safe cofferdam requires a seal that is a minimum of 9.6' thick plus a safety factor. The safety factor can be additional concrete thickness, reliable concrete-to-HP bonding, or a combination of the two.

This subject is important, keep working with it or asking questions as needed.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[Guest090822]

I've attached a spreadsheet with all 3 approaches.

I think I get it now.

Thanks again!

 

[SlideRuleEra]

Edit:

Ok, I see you are using the submerged weight of concrete (77.6 PCF). In that case I agree with the math for D = 14.5' for concrete only and D = 6.57' for concrete plus concrete-to-HP bond strength.

Even so, I would not go with a seal less than 10' thick... and that takes credit for a certain amount of bond strength.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[Guest090822]

I've attached the guidance I used.

 

[Okiryu]

Just a question related to this thread. Are there any limitations for large water heads when using concrete tremie seals? I guess that for this thread, for a water head of 12', a concrete tremie I practical. What can be recommended for deep waters? Say about 45' of water head...soil-cement improvements (deep soil mixing) at the bottom of the cofferdam?

 

[SlideRuleEra]

Okiryu - The deepest cofferdam that I have worked on was about 30' and I believe used a concrete tremie seal. That was long ago, while I was still learning the basics from my father, Don't recall the details. Like many in my experience, that one was in open water not on land. In open water, there are no issues with external soil pressure on the sheet piling... but there are different concerns. The most difficult are cofferdams that are part in water and part on land, i.e. waterfront pump stations at our electric generating stations.

One of the documents that I posted for the OP's review has this to say:

From a practical standpoint, cofferdams are limited to 60-foot long sheet piles. Manufacturing,
transporting, handling, threading and driving sheets longer than 60 feet creates major
problems. We have built cofferdams with 40 feet of dewatered depth by excavating below the
tip of 60-foot long sheet piles. This could be done only because the ground below the sheet
piles was stiff enough to stand vertically under water long enough to place the tremie concrete.
We have also chemically grouted sand lens to prevent underwater cave-ins. Usually tremie
sealed cofferdams are limited to about 30-feet or less of dewatered depth, plus a 30-foot deep
tremie seal.

I doubt soil-cement would work since a tremie seal is constructed underwater.

For really deep projects an open caisson may be used. See this write-up of one here in Charleston (poor soil, high water table) - Link

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[Okiryu]

Hi SRE, thanks for the reply. I have to check it but I believe that deep soil mixing improvements have been also conducted for marine applications in Japan. I will update later if I find something... thanks again for the input.

 

[PEinc]

If you take the unbalanced water head to the bottom of the tremie seal, yes, you can use the full weight of the concrete. However, the water head will be greater than the head you used with buoyant concrete for the tremie. If you try using the head to the bottom of the tremie, you will need to use an iterative design approach because you will need to guess the tremie thickness to get the head pressure.

And again I remind you to include the safety factor against uplift.

http://www.PeirceEngineering.com