I think that in addition to the initial moisture freezing (leading to surface cracking and spalling) you would get some significant thermal stresses due to the rapid change in temperature at the surface.
These stresses could go deep into the concrete and cause problems with deep cracking and strain.
How much liquid nitrogen are we talking about?
A small amount would boil off quite rapidly and not use as much heat in the concrete and not cause as much damage. A large amount that took some time (and heat) to boil off would do a lot more damage.
Is this a hypothetical question or did a spill occur?
Did it occur in a critical area or just on a non structural pad?
If you had a spill I would recommend some detailed investigation.
If the spill was on a non structural pad then perhaps replacing it or a simple overlay would be more economical than a detailed investigation. Simply doing nothing and waiting for any problems to show up might be a viable option for a non structural pad.
If it’s a structural member than a significant field and office investigation could be indicated.
Field investigation could consist of a mix of destructive (core samples) and non destructive testing (ultrasound and/or x-ray) as warranted by conditions.
You could do some worst case analysis and assume that the heat came completely from the concrete, and model the thermal stresses that could be created. This would most likely be some sort of finite element analysis knowing the thermal properties of the concrete (insulating properties, specific heat of the concrete and coefficients of thermal expansion) as well as the heat necessary to boil the nitrogen. If you know how long to boil the nitrogen, you could determine if the concrete was capable of giving off the heat fast enough to cause all the boiling or is it was contributing its heat at the maximum rate possible and adjust your model accordingly. This sort of analysis could be done if you are doing a risk analysis for a potential problem or actually had a spill.
More details of the problem would be helpful.
Rick Kitson MBA P.Eng
Construction Project Management
From conception to completion
(Thermal shock) There were visible signs of both surface cracks. The entire corner was cracked through the slab on the sides and would likely spall in time.
I do not know really how long the spill lasted. According to the client, a valve malfunctioned and remained in the open position – completely hydrating the pad (or at least the corner under the valve). He did not say how long the valve remained open, but given the location, I suspect it may have been awhile before it was noticed. I do not if the tank is monitored or alarmed for such malfunctions and can only assume it was not. My guess is that someone noticed the cloud produced.
This is a real occurrence.
The pad is structural the spill occured in a critical area. The tank is supported by three legs. The spill was directly under one of the legs. The tank weighs 18,000 lbs when operational.
Upon recommendation from a colleage at ICRI, we decided to take 3 cores samples and do strength testing. It was also recommended to perform the ultrasound/x-ray tests, but we decided to see what came of the first test.
We subcontracted a concrete testing company to come out and take core samples. They were able to get one core right in the middle of the spill zone. They drilled down 18” without a problem. The problems started when they tried to remove the core sample. It broke off 6” from the top. They attempted to get the rest of the core out (by wedging a screwdriver between the core and the slab and trying to lift it out.) to no avail. The concrete kept breaking into pieces.
I have the top 6” of the core and the pieces. You can easily crush the smaller pieces they chipped out. The partial core has visible cracks all through it.
If you tapped a hammer on the spill area concrete, it gave back a hollow sound. As you continued to tap away from the spill area, the sound eventually changed from hollow to solid. I know this is not very scientific, but it told us that the damage did not necessarily extend throughout the entire slab (8’ x 8’ x 18” approx)
It was out intent to take 2 more samples moving away from the spill for comparison, but we were limited by the clearance under the tank and the contractor’s equipment. I would be interested to see how far the cracking extended into the pad.
There is no available data on the pad. It was determined at the time of the core sample, that the pad had 2 mats of #5 bars, the top one being 6” below the surface. From the areas exposed above grade on the side of the pad, it looked like there was a cold joint in the pad. We have no way of knowing if there was any past modifications done to the pad or if the second truck was late in arriving at the site for the pour. All we have is speculation.
The end result:
We recommended full replacement of the pad. Without having any real data, other than observations, and the structural function of the pad, we felt this was the best option.
The client was hoping for only a partial replacement. In order to do that, the tank will have to be removed or supported independent of the pad and the extent of the damage will have to be determined in the field (or by taking the additional cores and performing the strength tests). It is more cost effective to go ahead and replace the entire pad.
The report, recommendations and design (for a new pad) are in the hands of the client and the construction company hired to make the repair/replacement. I plan to be on the site when they start to “see” more of the damage and get a better understanding of how far it migrated through the pad.
Based on your description of the extent of the damage, I agree with your recommendation to remove and replace the pad. Its often more economical and always surer to completely remove and replace a problem than to do a lot testing and examination, which may result in the same end result. Even a partial replacement would cause some disruption to operations, a full removal may be quicker.
In your case you do not have sufficient knowledge of the construction of the pad to design a patch until the demolition is complete thus causing further delay.
I never like doing a partial removal and replacement of a structural member. You then have to be concerned with the quality of the patch. In concrete how sure will you be that the surface is adequately prepared, appropriate bonding agents and admixtures used etc? In a total removal and replacement quality control is easier.
The shear failure in the core sample could be a number of things. Sometimes cores shear due to vibration in the drilling. It could be that the sudden contraction in the surface of the concrete caused the failure. It could be the cold joint in the pour that you mentioned. This could also explain the hollow sound when tapping.
With all the unknowns, a complete removal and replacement is the way to go.
Let us know how the fix works out and what additional information you can gather during the demolition.
If your firm has an instrumentation division, you might want to sell a monitoring system that would at least identify this type of leak. At least make the recommendation to the client that the equipment be monitored.
Rick Kitson MBA P.Eng
Construction Project Management
From conception to completion
