Open Access
Miller, David Matthew
Graduate Program:
Industrial Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
Committee Members:
  • Edward Demeter, Thesis Advisor
  • underwater structural adhesives
  • photocurable acrylic adhesives
  • adhesive aging
The vast use of polymeric substances in marine applications has spawned many caulks, paints, adhesives, and composite materials capable of resisting a wide range of extreme environmental conditions. Many of the underwater caulks and sealants are not intended for load bearing, as they are extremely elastic and inherently weak and therefore would not suitable as an underwater structural adhesive. These materials are commonly cured by adding a secondary chemical component that acts a polymerization catalyst, such as in a two-part marine epoxy paint system or two-part epoxy adhesive, but they can take days to fully cure and must be applied to dry-docked vessels. Another method of curing a specific type of polymer is the use of light in the ultraviolet and/or lower visible light region of the electromagnetic spectrum. This is known as photocuring, and the most profound advantage for use of this cure mechanism is the ability to create an optimally cured adhesive joint in approximately 30-60 seconds. Photocurable acrylic adhesives contain organic or inorganic photoinitiators that split into free-radicals that polymerize methacrylate monomers. Developed as a novel approach to workholding and an alternative to mechanical, pneumatic, or hydraulic fixturing mechanisms, Light Activated Adhesive Gripper (LAAG) utilizes a photocurable acrylic adhesive as the main component of the system. The preferred adhesive for bonding to ferrous materials is CTECH® 14-90-2. The patented concept employs sapphire pins imbedded in a structural housing to allow for the transmission of UV light into the adhesive that subsequently adheres to the work-piece. The number of discrete adhesive joints, the size of the sapphire pins, and the orientation of the pins can be optimally designed using specialized Finite Element Analysis software. The preferred light source is a metal halide UV spot curing lamp. LAAG fixtures have been proven to exhibit maximum dynamic rigidity and high holding strengths, and the system allows for maximum tool access to the work-piece. Previous testing has confirmed tensile strengths approaching 55 MPa which makes the hardware and concepts of LAAG technology a viable candidate for underwater structural repair. Much characterization work has been performed with CTECH® 14-90-2 pertaining to LAAG technology, including the characterization of optimal cure rates and critical irradiance values, but there has been no investigation of its ability to adhere to submerged substrates or the resistance to extreme environmental conditions. To begin this investigation, the performance of the adhesive when bonded to dry substrates (dry assembly) was evaluated. This investigation revealed that CTECH® 14-90-2 adequately wets both smooth (ground) and rough (bead blasted) when dry assembled. The means and the variations were barely statistically equal at 95% confidence level with a F-test p-value of 0.126 and a T-test p-value of 0.067, respectively. Also, the sensitivity of dry assembled CTECH® 14-90-2 to temperature was characterized. The means of the axial load strength of the adhesive joints at various temperatures were as follows: 3141N at ~0º F, 2225N at ~38º F, 1535N at ~70º F, 1171N at ~96º F, and 763N at ~120º F. It is evident that increasing temperatures resulted in lower strengths of the adhesive joints, resulting from a decreased dynamic modulus. The inverse linear relationship was confirmed by regression analysis using Minitab® and the model fit the data quite well with a R2 value of 91.5%. The sensitivity of the strength of CTECH® 14-90-2 to thermal excursion was evaluated. The samples were allowed to thermally soak to either 0 ºF or 120 ºF for eight hours and then thermally soak back to room temperature. The statistical analysis confirmed that the 0 ºF or 120 ºF excursions resulted in more variation when compared to a room temperature control sample. The F-test p-values were 0.000 and 0.012, respectively. In both cases the estimates for variance was double the variance of data without an excursion. It was also statistically proven (T-test p-value of 0.004) that the very hot (120 ºF) excursion sample was taken from a distribution with a higher mean (estimated at 231N) with respect to the very cold excursion and the room control means. The means of the very cold excursion and the room control were concluded to be equal (T-test p-value of 0.127). This phenomenon is believed to be attributed to the interaction between dark polymerization (continued polymerization after the UV light source has been removed) and the thermal excursions. The ability of CTECH® 14-90-2 to wet and adhere to submerged substrates (wet assembly) was evaluated. This investigation revealed that wet assembled CTECH® 14-90-2 joints are inherently (estimated 400 N) weaker than dry assembled joints when bonded to smooth (ground) substrates (T-test p-value of 0.009). The joints were further reduced when wet assembling to rough (bead blasted) substrates (T-test p-value of 0.000). These results were attributed to inadequate wetting of the substrate due to entrapped water in the random depressions left by bead blasting. The effect of saltwater exposure on uncured adhesive at various temperatures was also studied. An ANOVA was performed using the experimental data. It was concluded that exposure time was a significant factor (F-test p-value of 0.013). Temperature and the interaction were not significant (F-test p-values of 0.948 and 0.514, respectively). The longer that the uncured adhesive was exposed to saltwater the lower the subsequent axial load strength was. An investigation revealed that temperature and time were significant factors with respect to de-lamination (hydrolysis) of CTECH® 14-90-2 adhesive joints under saltwater submersion. Increasing ambient saltwater temperatures resulted in faster hydrolysis rates. Hydrolysis occured first at the sapphire/adhesive interface, as CTECH® 14-90-2 continued to adhere to the steel substrates. Statistical comparison of means were conducted to aid in the determination of when the adhesive joint significantly decreased in strength. When there was at least 95% confidence that the mean was reduced from the initial immediate pull strength, it was concluded that there was a significant drop in strength. Wet assembled joint strengths in hot (96 ºF) water were significantly reduced after three days, while dry assembled joint strengths took six days to reduce. Recall that wet assembled joints are inherently weaker than dry assembled joints. In room temperature saltwater, the joint strengths were not reduced until twelve days of submersion. The reduction in joint strength was attributed to hydrolysis in all cases. In cold temperature saltwater, hydrolysis was stymied for at least forty-eight days. Five chemical components were believed to affect hydrolysis rate: impact modifier, reinforcing agent, acid #1, acid #2, and acid #3. The exact chemical details of these chemical components are intentionally forgone due to proprietary reasons. A fractional experimental design was conducted with assistance from CTECH®. An initial round of experiments led to valuable insight on what chemical components were needed for immediate pull strength. An ANOVA was conducted on the experimental data. It was concluded that the impact modifier and the reinforcing agent have the most profound affect on the immediate pull tensile strength. The impact modifier and reinforcing agent had F-test p-values of 0.000 implying a strong significance. The acids #1, #2, and #3 also proved to be significant with confidence levels of 0.016, 0.013, and 0.031, respectively. The main effects and interaction plots showed that the impact modifier and reinforcing agent must be at high levels for the maximum immediate pull strength. The second round of experiments led to valuable insight on the chemical component’s affect on the rate of hydrolysis. It was found that acid #1 and acid #3 should be removed from the formulation to attain the best resistance to hydrolysis (F-test p-values of 0.000). Of the sixteen test adhesives that were formulated, several had immediate pull strengths similar to that of the current CTECH® 14-90-2 formulation. Several adhesives also have shown an increase in resistance to hydrolysis in hot water conditions, but CTECH® Test Adhesive #3 proved to have the best results. CTECH® Test Adhesive #3 had an average dry assembly, immediate pull strength of 1478N, and after twelve days of hot water storage the residual strength was 908N. Although hydrolysis is still present when using CTECH Test Adhesive #3, the detrimental impact has been reduced when compared to the 198N average when the current CTECH® 14-90-2 formulation underwent the same submersion experiment. This was attributed to the absence of acid#1 and acid #3 in CTECH® Test Adhesive #3. CTECH® Test Adhesive #8 was chemically opposite to CTECH® Test Adhesive #3 with respect to acid#1 and acid #3. CTECH® Test Adhesive #8 completely hydrolyzed after twelve days of submersion in hot saltwater. It averaged an immediate pull strength 1183N. An investigation of the use of silane primer coated on both substrates led to the conclusion that silane coated substrates will result in higher wet assembly strengths, but silane did not cause any resistance to hydrolysis. After extended investigations, CTECH® Test Adhesive #3 proved to have the best resistance to hydrolysis and was adequate for underwater construction in hot temperature (96°F) saltwater conditions for at least twelve days, assuming that an average of 800 N is adequate. The same was true for underwater construction in cold temperature (38°F) and room temperature (70°F) for at least forty-eight days. An ad-hoc experiment was performed to test whether CTECH® Test Adhesive #3 would adhere to polymeric epoxy based marine anti-corrosion paint. CTECH® Test Adhesive #3 averaged 777N under hot water conditions for twelve days, but most importantly the adhesive tore away paint when destroyed after the test duration expired, proving that the adhesive and the interfaces were stronger than the paint. Also, CTECH® Test Adhesive #3 performed equal to the current CTECH® 14-90-2 formulation when wet assembled. It had an average value of 1096N when wet assembled and immediately pulled. When wet assembling and stored in hot temperature salt water for twelve days, CTECH® Test Adhesive #3 averaged a value of 417N when the current formulation of CTECH® 14-90-2 averaged 12N. Several chassis were fabricated using a plexi-glass pin instead of a sapphire pin. These were bonded to plexi-glass anvils with CTECH® 14-90-2 and were stored for twelve days in hot saltwater (96°F). Although some samples exhibited no hydrolysis when destroyed, others had significant hydrolysis. The results were very erratic and this was the major cause for caution when using plexi-glass pins. This was attributed to the fact that the output light level was 50% that of all previous experiments, so that the plexi-glass did not melt. It was possible that the joints were not receiving the proper dosage of light to cure the adhesive. CTECH® #3 was bonded to plexi-glass and there was also rapid hydrolysis. Lastly, CTECH® #8 is chemically opposite to CTECH® #3 with respect to acid #1 and acid #3 so it was suspected that rapid hydrolysis would occur. Rapid hydrolysis was confirmed after only twelve hours. The hydrolysis was mainly evident on the sapphire pin, but not on the steel substrate.