Mixed Mode Fracture Characterization of Hydrogel-3D Printed Polymer Adhesively Bonded Systems 

By Nicholas Sardinha

Introduction 

Hydrogels are soft polymer networks with an interstitial liquid, usually water, and they have many applications ranging from tissue engineering, water purification [1], and even anti- biofouling. However, for hydrogels to be more useful for anti-biofouling, their adhesion to different surfaces must be investigated. Many marine vessels have hulls made of different metals such as steel and aluminum, but with the rise of additive manufacturing came the increase in the use of 3D-printed polymers as hull materials. This research attempted to study the adhesive bonding of hydrogel and 3D-printed acrylonitrile butadiene styrene (ABS). The specific method chosen to evaluate the fracture toughness was the J-integral method, used on a mixed mode bending fixture described in the methodology section of this report and based on the work of Zhao Y, Seah LK, Chai GB [2]. Mixed mode bending applies opening mode-I and shearing mode-II to the double cantilever beam (DCB) specimen. However, this type of test has never been conducted with two materials with such a large difference in elastic moduli, with hydrogel having a Young’s modulus in the 10 kPa – 2 MPa range and ABS in the 2-4 GPa range. This causes the ABS to experience much less deformation than the hydrogel at the same load. That is the primary reason why the traditional peel test was not used, as hydrogel experiences significant deformation, which absorbs energy. This leads to the peel test not accurately measuring fracture toughness of softer materials, since that additional energy absorption is not properly accounted for in fracture toughness calculations. 

Methodology 

Specimen Production 

Table 1: Hydrogel Composition 

Number of Samples	Total Solution (ml)	DI Water (ml)	AAm (g)	DAC (ml)	MBAA  (mg)	KA (mg)
1	25	14.8	8.9	10.2	23.2	13.7
3	75	44.4	26.7	30.6	69.6	41.1

The specimens used in this research were fabricated using a unique in situ fabrication method. Firstly, a hydrogel precursor solution was made using the five main components: deionized water (DI Water), Acrylamide (AAm) monomer, Acryloyloxyethyltrimethyl ammonium chloride (DAC) cationic monomer, N,N’-Methylenebisacrylamide (MBAA) crosslinker, and α-Ketoglutaric acid (KA) photoinitiator. From Table 1 the weight percentage of AAm and DAC are 25.24 wt.% and 32.69 wt.%, respectively. This process produces a clear hydrogel precursor solution that can be used for the fabrication of test specimens. 

The second half of each specimen is a 3D-printed ABS T-shape, which was produced using a Stratasys uPrint 3D printer and ABS P430. After the ABS T-shape finished printing, a 10 mm long pre-crack film was glued to the top surface, as shown in the red area of Figure 1(a). 

 For the control samples, the rest of the 3D-printed surface was left unaltered, and the ABS T- shape was placed into a silicone mold. This silicone mold was designed with two connected cavities: one for the ABS and the other for the hydrogel precursor solution. The side for the precursor solution had an opening that was sealed using a strip of photocopier film and two glass microscope slides. The purpose of using transparent materials to seal that side of the mold is to ensure that the 365 nm ultraviolet light can pass through and photocure the hydrogel. After many experiments, a 4-hour photocuring time was chosen and used for all samples. Since the mold had an opening on the top for the hydrogel precursor solution to be poured in, a sheet of plastic wrap was applied around the entire mold to prevent moisture from entering or exiting during curing. Photocopier film was also used to line the inside of the mold to ensure that the surface of the hydrogel would be smoother. 

After the samples finished curing, they were removed from the mold, and excess hydrogel was cut away with a razor blade so that the edge of the hydrogel aligned with the edge of the ABS. Once this was completed, a 20× 100 mm2 strip of photocopier film was superglued to the backside of the hydrogel. Next, a 100 mm long strip of paper with millimeter markings was applied to the side of the ABS to help measure the pre-crack length. To ensure that the pre- crack area did not have any adhesion, the crack was slightly opened by hand, and the end of the pre-crack was marked on the side of the ABS with a black marker. 

Photooxidation 

The photooxidation process refers to the exposure of the ABS surface to chlorine dioxide (ClO2) gas and 365 nm ultraviolet light. To produce ClO2, 100 mg of sodium chlorite (NaClO2) was added to 14 ml of DI water inside a small dish. This small dish was then transferred into a larger glass reaction vessel. Any ABS samples to be photooxidized were then placed in the reaction vessel close to the small dish. The entire reaction vessel was then placed in the fume hood. Then using an adjustable pipette, 100 µl of 37% hydrochloric acid (HCl) was added to the solution, which starts producing chlorine dioxide gas. After that, the reaction vessel was sealed using plastic wrap and a clear flat plastic sheet to contain all the produced gas. Then the ultraviolet light was placed on top of the reaction vessel, and a timer was started as soon as the light was switched on. After the desired time was completed, the ABS samples were transferred into another container, where they could be used immediately for either contact angle measurements or for fabrication of mixed mode bending specimens. All remaining reacted solution was then transferred into a designated waste container specifically for that solution. 

Mixed Mode Bending Fixture 

Figure 1:(a) Hydrogel/ABS adhesive system and (b) T-Shape drawing with dimensions in millimeters. 

 

Figure 1(a) shows the ABS/hydrogel adhesive systems, and Figure 1(b) shows the dimensions of the T-shape part of either portion. The double T-shape specimens were placed into a mixed-mode bending fixture as shown below in Figure 2. This fixture was loaded under compression with a load P, where force and displacement measurements were recorded using a Shimadzu AGS-X universal testing machine (UTM) with a 50N load cell. A Celestron handheld digital microscope was used to capture the crack initiation. The loads at locations A, B, C, and D were determined using statics, and the angles measured at these four locations were used in determining the fracture toughness. Due to the significant difference in elastic moduli between the hydrogel and ABS, θB & θD are extremely small and approximated as zero. 

 

Figure 2: Mixed mode bending loading configuration of the adhesive system. 

 

The lever length c can be adjusted to change the mode mixity, and for these experiments a length of 30.61 mm was used. The horizontal distance between points B and C called length L in Figure 2 cannot be adjusted and was set to 45 mm. Prior to each set of tests, the load 50 N load cell was calibrated. 

To secure the prepped samples into the fixture, superglue was applied to the top surface of the hydrogel to meet the grip, and the grips were tightened down. The lower grips on the ABS side did not require superglue, since the ABS did not deform during testing. Once the specimen was in the fixture, the pivot was slowly lowered until it contacted the surface, then a vertical line was drawn onto the hydrogel to measure the angle at point C. Both the force and displacement were then zeroed. A loading rate of 4 mm per minute was used based on previous mixed mode tests. The recording on the digital microscope and the UTM were started at the same time and two screenshots were taken, one at the start of the experiment and one at fracture initiation. 

These screenshots were then imported into MATLAB, where a code was used to analyze the angles of 3 different points. The first screenshot was analyzed to measure the initial angle of the vertical line at point C, and the second screenshot was analyzed to measure the angle of the fixture, the angle at A, and the angle at C (this is shown below in Figure 3). The time of fracture initiation was also recorded from this video and was used to get the load and displacement at fracture initiation. With the P, θA, & θC, the fracture toughness can be calculated using the following formula:

= 100( sin() + sin() − sin() + sin() ) 

The average fracture toughness and 95% confidence interval were plotted using Excel. The force vs displacement was also plotted using MATLAB and the Excel spreadsheets produced by the UTM. After the completion of each test, the hydrogel was separated from the ABS and the ABS surface was cleaned using deionized water for future use. This procedure was repeated until four consistent tests were performed. 

Figure 3: Example of Angle Measurements at Fracture Initiation. 

 

Water Contact Angle Measurement 

Water contact angle measurements were taken using a Ramé-Hart Model 90 goniometer. The goniometer was calibrated using the included calibration sphere and the water contact angle software package. The surface to be tested was placed on the test platform with the backlight turned on and leveled using the included adjustment knobs. After that, a droplet of deionized water was dropped on the surface using the included syringe, and the contact angle was recorded. This was repeated in four separate locations on the ABS surface on two different specimens. This procedure was repeated for the three different conditions which were untreated, 15-minute photooxidation, and 30-minute photooxidation. The average and 95% confidence interval were then plotted in Excel. 

Results 

Figure 4: Water Contact Angle for Untreated, 15 Minute Photooxidation, and 30 Minute Photooxidation. 

 

The water contact angle measurements from Figure 4 showed a significant difference between untreated samples and samples that have had the photooxidation procedure done on them. The average contact angle decreased from 76.23° to 42.06°. This was an expected result as the photooxidation process modifies the ABS surface with more reactive groups, meaning the surface should become more hydrophilic, which it did. It also showed that increasing the photooxidation treatment time from 15 minutes to 30 minutes did not have much of an effect. This is important information as it led to one of the challenges experienced later on in the process. 

 

Figure 5: Control Load vs Displacement. 

 

The load versus displacement graph in Figure 5 shows the same trend for all specimens. There is an initial shallower slope section that quickly transitions into a linear portion until the peak load is achieved, and then the load quickly drops off as the crack grows along the interface. All four specimens achieved peak load values within 1 N of each other. 

Figure 6: Control Fracture Initiation Toughness. 

 

The average fracture toughness for the untreated control samples was 31.1 J/m2 as seen in Figure 6. There are no other conditions to compare to the control which is explained in the discussion section of this report. 

Discussion 

Many challenges were faced during this research that limited the scope of this project. Due to the novelty of bonding hydrogel and ABS plastic, a unique hydrogel precursor solution had to be created and iterated on. Initially, the hydrogel used had half of the concentration of AAm and DAC, which caused the hydrogel to be very soft, excessively deform, and ultimately lead to inconclusive tests. This challenge was eventually overcome by doubling the concentration of AAm and DAC, along with increasing the MBAA cross-linker concentration. 

This led to very successful control experiments. However, after the control tests concluded, photooxidation tests started. Based on the contact angle results, a 10-minute photooxidation procedure was selected. As expected, the photooxidation had a major effect on the adhesive bond between the surfaces; it had such a large effect that the hydrogel consistently failed prior to the interfacial bond. Corrective measures were attempted, such as decreasing the concentration of the NaClO2 and HCL solution by half and the exposure time down to 3 minutes. This still resulted in the same challenges as previous tests, namely crack propagation into the hydrogel. 

The second-largest challenge was determining the proper method to measure θC. The methods used by other researchers who used the same fixture did not work as expected due to the ABS acting as a rigid body and causing the hydrogel to indent rather than bend. This indentation issue was compounded by the fact that the hydrogel used in these experiments was prone to warping. Many different solutions were proposed to combat those problems, but ultimately, applying a backing layer to the hydrogel and drawing a vertical line at point C was chosen, as described in the methodology section of this report. 

Conclusion / Future Work 

There are many more aspects of this research topic that can be further investigated, the most pressing being the degree to which the photooxidation process increases the fracture toughness. This would require the control samples to be redone using a hydrogel stronger than an interfacial bond to prevent cracks from propagating into the hydrogel. Once the hydrogel is strong enough to withstand the photo-oxidized tests, other test configurations can be tested, such as modifying the surface geometry or roughness. Following that, another future step for this research is using digital image correlation (DIC) to determine the mode mixity of these mix mode tests. The lever length used for these tests was chosen because, with the same material for top and bottom T-shapes, it produces equal parts mode I & II loading conditions. However, since the top and bottom are not the same material, the mixity is not the same. With the help of DIC, the strain fields near the crack tip can be analyzed to determine the true mode mixity. Overall, this research has built up some knowledge of hydrogel and ABS adhesively bonded systems. Even though it did not achieve all of its goals from its onset, it still provides a solid foundation that can be built upon by future researchers. 

References 

Yahia LH, Chirani N, Gritsch L, et al. History and Applications of Hydrogels. J Biomedical Sci. 2015, 4:2. http://dx.doi.org/10.4172/2254-609X.100013. 

 Y. Zhao, L. K. Seah, G. B. Chai, Measurement of interlaminar fracture properties of composites using the J-integral method, J. Reinf. Plast. Compos. 35 (2016) 1143-1154. https://doi.org/10.1177/0731684416642031.