This article describes the failure of coatings applied to the concrete floors of a county detention facility. The structure is a multistory building with modular cells and concrete floors on each inmate level. The main structural floors were constructed of lightweight aggregate and the mezzanine floors constructed of a normal-weight aggregate. The facility was relatively new, becoming operational in September of 2015. The building could accommodate 608 inmates with the ability to add additional space as the population demanded.
The coating failure, which revealed itself via delamination, cracking, heaving and crumbling, began within a few months of opening, and presented in both isolated and progressive failures on multiple levels but appeared to be predominantly present on the mezzanine floors. Repairs were originally performed using coating materials supplied by the original product supplier. Later, epoxy topping repairs employed materials from a different supplier for test applications. The failures continued, prompting this investigation.
The engineering firm elected to conduct the field investigation themselves and the investigator was provided with background and technical information, photographs and intelligence from the observations of others. Core samples representative of failing and non-failing floor surfaces were requested by the investigator. The core samples were removed from four locations selected by the client and included two failing and two non-failing samples. Since the samples were selected based on whether the coating was failing or non-failing, the failing samples were both removed from the same floor type (mezzanine) and the non-failing samples from the main structural floor. The possible influence of floor type was not considered.
As seen in JPCL, April 2016
THE INVESTIGATION
Not having the opportunity to conduct a site investigation can be a handicap as a great deal of information can be determined through that process. It is possible that some details of the surface preparation and coating application methods could’ve been discerned leading to reasonable hypotheses as to possible causes of the failure.
For example, testing for the presence of water beneath delaminated coating could’ve been conducted, as well as judging the difficulty of removing the coating from the substrate or separating coating layers in the field. Such information will generally provide more meaningful information than that evaluated in the laboratory. In fact, field findings are commonly the basis used for determining which initial laboratory examinations should be undertaken so that hypotheses as to the failure from the field investigation can then be tested through sample analysis.
REVIEWING THE AVAILABLE INFORMATION
Supplied background technical information coupled with site photographs assisted in the determination of a specific path to follow. An epoxy flooring system was originally specified consisting of four distinct coating layers: a primer at 6-to-8 mils, an aggregate-filled base coat to a nominal 250 mils (1/4 inch) a grout coat at 8-to-32 mils (to seal the base coat) and a topcoat at 6-to-8 mils. The intent of the work, per the project specification was “… to provide a seamless epoxy floor and base.” The system design also included:
- a waterproofing membrane (to be furnished by the manufacturer of the epoxy) that was a trowel-applied, reinforced latex composition strengthened with glass mat or plastic fabric (or equal);
- a specified aggregate; well mixed throughout the thickness of the flooring (not sprinkled or broadcast on top); and
- a transition coat of homogeneous methyl methacrylate (acrylic).
Surface preparation specs dictated preparing concrete surfaces by removal of laitance and other interference materials. Installation requirements consisted of:
- coating the floors and wall bases with troweled adhesion waterproofing;
- embedding the reinforcing fabric;
- trowel- or brush-applying additional waterproofing coating to seal voids;
- applying a thin coat of the manufacturer’s standard bond coat; and
- trowel-applying the resin and aggregate mixture at a thickness of 1/4-inch.
PHOTOGRAPHIC EVIDENCE
Numerous photographs were provided showing the nature and distribution of the detention facility floor coatings. Because moisture is a common problem with concrete surfaces the photographs were reviewed for evidence of this possible cause (Fig. 1).
Fig. 1: This photo demonstrates that not only are there cracks and delamination present, but it provides clear evidence of water intrusion. The visible rust stains are only found when ferrous metals begin to corrode. All figures courtesy of the author |
Fig. 2: It is imperative to locate the source of even small leaks as they can often lead to significant moisture issues within a coating system. |
When free water is contributing to a coating failure it is important not only to demonstrate this through the evidence but also to seek out the source. The engineering firm was encouraged to look for likely sources of free water and one source not to be ignored is floor cleaning. Since cracks were present it is possible that cleaning water seeped through cracks to the substrate. While this theory could not be ruled out, Figure 2 shows that even small leaks can be significant sources of water.
In addition to the evidence confirming that water was contributing to the coating failure, another likely contributing factor was observed in the photographs — the crumbling of a thick monolithic cementitious layer, shown in Figure 3. The integrity of a concrete layer and the aggregate-filled epoxy coating layer is clearly compromised. This issue could be related to the concrete itself being a non-durable substrate, or the result of a coating issue related to the aggregate-filled coating material. Cracks in the coating can result from moisture vapor pressure contributing to coating delamination. Rigid aggregate-filled films do not recover from deformation. They are subject to flexure when loose which induces cracking. A site visit could have served to clarify the initiating factor.
Fig. 3: The failure mechanism also suggests areas without a durable substrate. |
Fig. 4: This photo shows evidence of standing water which has stained the cell floor. Cracks in the coping around the base of the walls allow the water to enter. |
CORE SAMPLES
The four concrete core samples requested were provided to the investigator. The two from failing areas and two from non-failing areas (used as control samples) are described in Table 1.
Table 1: Concrete Core Floor and Coating Samples
Non-Failing Samples | Failing Samples | ||||
Sample | Location | Floor Type | Sample | Location | Floor Type |
Core Specimen #1 | Level 10 South Cell 10B05 | Main Structural Floor Light Weight Aggregate | Core Specimen #2 | Level 10 South Cell 10A44 | Mezzanine Normal Weight Aggregate |
Core Specimen #4 | Level 6 North Cell 6D01 | Main Structural Floor Light Weight Aggregate | Core Specimen #3 | Level 8 North Shower | Mezzanine Normal Weight Aggregate |
VISUAL AND MICROSCOPIC EXAMINATION
In addition to visual inspection, microscopic examination of the samples was conducted using a digital microscope with a magnification to 200X. Findings from the handling and examination of the cores follow.
Sample 1: Non-Failing Core Sample Findings
- The coating could be pried from the edges of the core and delamination could be forced with gentle pressure.
- The plane of failure appeared to exist between the layer with the colored aggregate and another layer with a translucent coating and clear aggregate that remained tightly adhered to the core sample.
- The colored aggregate layer showed encapsulation by resinous material when the cross-section was viewed.
Sample 2: Failing Core Sample Findings
- The sample arrived with coating material completely delaminated from the concrete substrate.
- The plane of separation (failure) was between the colored aggregate layer and a translucent layer that was present on the concrete.
- In one small area the translucent layer was noted as present on the back of the delaminated chip.
- The clear aggregate associated with the translucent layer was significantly less dense on this sample as compared to Sample 1.
- The colored aggregate appeared more densely packed in this sample as opposed to Sample 1 with less encapsulation by the resinous material.
- The surface of the substrate was viewed and revealed the translucent material as well as a white powdery material that was likely from the resin of the colored aggregate layer.
Sample 3: Failing Core Sample Findings
- The coating material was delaminated from the top surface upon receipt by the laboratory.
- The plane of failure was between the colored aggregate and the translucent and fibrous material that appeared to be a surface layer over the concrete.
- The top surface of the delaminating coating chip had many cracks in the top surface of the clear coat.
- In some areas the colored aggregate appeared to be protruding through the clear coat.
- The bottom surface consisted primarily of the colored aggregate but also showed a white, powdery residue present between the spaces and along the aggregate, but did not encapsulate the aggregate.
- The colored aggregate could be loosened by gentle scraping.
- The colored aggregate was observed in the clear topcoat, suggesting that it was loose at the time of the clear coat application.
- There was minimal resinous material present in the layer filled with colored aggregate.
- The area under the delaminated coating showed fibers, a translucent material, and a material that was cementitious. Voids were noted through this cross-section.
Sample 4: Non-Failing Core Sample Findings
- The coating material could be lifted and delaminated from the edges of the core using moderate pressure from the fingertips.
- The plane of failure appeared to be between the colored aggregate layer and a clear aggregate with a possible translucent (primer) resin binding it to the concrete surface.
- When probed with a knife, the clear aggregate could not be easily removed, but the colored aggregate could be loosened.
- The colorfed aggregate in this sample could not be loosened as easily as with previous samples. In one area the colored aggregate layer left a spot with aggregate and a powdery residue on the concrete core when removed.
- The top surface of the coating material was viewed and a few cracks were noted through the topcoat. The aggregate was completely covered by the topcoat and was not protruding through the top surface.
- The bottom surface of the delaminating coating material consisted of a colored aggregate material and appeared to have some resinous material present. This sample was significantly less powdery than previously examined samples where the aggregate was more loosely adhered.
- The area under the delaminated coating material left behind from the filled coating layer was some colored aggregate and some powdery residue.
- The clear aggregate did not appear to be encapsulated by the translucent material as seen in Sample 1. The colored-aggregate layer showed an increased presence of binder material around the colored aggregate.
- In addition, a waterproofing membrane system (with imbedded fiber) was called for in the specification but only found in Sample 3. There is no indication as to why the waterproofing membrane is absent at the other core locations.
INFRARED SPECTROSCOPY
Infrared spectroscopic analysis was performed in the lab. This technique involved combining sample scrapings with potassium bromide powder and forming pellets under high pressure. The pellets were then placed in the optical path of the spectrometer and spectra were obtained over the range of 4,000-to-400 cm-1. The clear topcoats from Samples 1 through 4 were consistent with epoxy resin.
The white powdery residue attributed to the layer filled with colored aggregate was consistent with crystalline silica. No evidence of resinous material was indicated by the spectrum produced by the powdery residue, which was present at the failure interface.
The translucent material on the bottom surface of Sample 2 was consistent with an epoxy resin. The presence of calcium carbonate was also indicated.
No methyl methacrylate (acrylic) coating layer was found.
SUMMARY
The coating system applied to the concrete of this facility was not the one specified. A four-coat system was specified and only two distinct coating layers were present in the core samples. The waterproofing membrane specified was only found to be present in one of the four core samples and although only one aggregate was specified, two (colored and clear) were found to be present. A plane of weak intercoat adhesion was noted in the field as delamination was found between a coating layer filled with colored aggregate and the subsequent layer on each of the samples submitted. Two concrete core samples from non-failing locations arrived with the coating system intact, but it easily separated with mild pressure. The two concrete core samples from failing locations arrived with the coating system separated.
Some of the coating failures were associated with moisture intrusion as evidenced by delamination with rust staining and efflorescence in coating cracks visible in the photographs provided (Fig. 3). It was reported that common plumbing chases between adjacent cells include water and waste connections. Leaks in these locations provide a path for water to travel along the concrete to the underside of the floor coating system. Once the moisture entered the concrete layer it migrated over a large area before showing evidence of its presence.
There was also evidence that underlying layers, cementitious and aggregate-filled, had crumbled along delaminated edges. This was a progressive deterioration most likely brought about by loading on a non-adhered edge when there was flexure of the delaminated layer. There were insufficient layers and the layers present were not at the appropriate thickness or consistency of binder-to-aggregate ratio, possibly due to mixing differences from location to location. This is supported by the failure of the aggregate to fully integrate into the binder.
FAILURE CONCLUSION
This case from the F-Files was presented to address the clues uncovered through examination of contract documents, information from on-site personnel, photographs depicting the condition of the coating and evidence of failure, and examination of coating samples, as there was no opportunity to visit the project site to conduct a hands-on examination. However, it was still possible to develop an understanding of what had occurred in the field and verify those theories through laboratory testing determining that the initial coating application, binder and aggregate mixing were flawed, the required film thickness was not achieved and the specified coating system had not been installed. It is not clear whether or not a substitution of materials was permitted. It is clear, however, that the system as installed was faulty. Add moisture to the equation and this flooring system was doomed to failure.
About the Author:
Rich Burgess is a senior consultant for KTA-Tator, Inc. where he has been employed for over 23 years. He is a member of SSPC and NACE and an active committee member for joint standards. Burgess is an SSPC-Certified Protective Coatings Specialist, a NACE-Certified Coating Inspector Level 3 (Peer Review) and an SSPC C-3 Supervisor/Competent Person for Deleading of Industrial Structures. In his current position, he performs coatings evaluations, coating failure analysis, specification preparation, expert witness and project management services for clients in the transportation, power generation, water/wastewater, shipping, marine and aerospace industries. Burgess is a principal instructor for the SSPC C-1, C-3, and C-5 courses, for the NACE CIP Program and a variety of KTA-offered training seminars. He holds a Bachelor of Science degree in Environmental Science from Rutgers University and a Master of Science in Operations Management from the University of Arkansas.
Reading this investigation/report, I agree 100% with the ‘Failure Conclusion’
I was just wandering if any water vapor moisture testing was performed prior to installing the coating system. But I agree with the fact the system was installed incorrectly and the possibility it was the wrong system if there was no substitution allowed.
Robert, I’m going to contact the author & I’ll let you know the answer as soon as I get it. Thanks, Jason -KTA Marketing Dpt.
Robert,
Thanks for your question and interest in KTA University. Unfortunately no moisture vapor (or concrete moisture content) data was provided (or perhaps conducted- we don’t know) but it could certainly contribute to the problem if indeed they were higher than recommended.
One significant drawback to not having the opportunity to visit the job site is many clues can go unnoticed by others.
I’m sure you’ve come across projects where the specification says “No Substitutions”
But they are made anyway, either by request and permitting or some other reason. It is also frowned upon when public (or quasi-public) contain requirements that do not allow for some mechanism to permit substitutions if equivalencies are available.
No paper trail was available to KTA to make a determination whether a substitution was allowd.
Regards,
Rich Burgess