4.1.

Mechanical Properties of the Bovine Artery

Stroke ranks as the third leading cause of death and the leading cause of long-term disability. The prevalent cause for stroke is the rupture of unstable atherosclerotic plaque within the sinus of the carotid artery that provides blood to the brain. To reduce the risk of stroke, it is essential to understand the mechanical conditions and consequent stress distribution in arterial walls that facilitate plaque rupture. As a first step, we adopted the use of DIC to determine the mechanical properties of the arterial wall.^{25 26 27 28} Previous studies on arterial tissues have used the “marker method,” which utilizes the change in location of three or four discrete marker points to extract the in-plane strain.^{29 30} One outstanding drawback is that the mechanical response of the tissue becomes homogenized over the marker grid. Arterial tissue is composed of layers of elastin and collagen and there is a high degree of nonlinearity in response to large strain.³¹ Local variations in the anisotropy and inhomogeniety of arterial tissues have not been examined in detail, partly due to the limitations of traditional experimental approaches.

In this preliminary study, bovine aorta and descending artery (total length from aortic arch of approx. 20 cm) were obtained within 1/2 h post mortem and maintained in saline solution. The arteries were maintained in an ethylene glycol tetra-acetic acid (EGTA) enhanced passive physiologic solution except during transport, dissection and testing. Each artery was cut axially between the thoracic ostia and laid flat, intimal side up. Rectangular specimens were excised, oriented axially, measuring approximately 25×100 mm. Specimens with a uniform thickness of approximately 2.0 mm were removed using a microtome blade, yielding arterial sections comprised of the intima and medial layers. The methodology used for specimen preparation is consistent with that used by previous investigators.³¹ Sectioning of the arterial samples using the microtome may have influenced the structural behavior of the tissue; damage to the tissue may have resulted from the sectioning process but was ignored. Prior to loading, each specimen was taken out of the EGTA solution and excess moisture on the intimal surface was wiped gently with a paper tissue. As the intimal surface is matte white, a peripheral coat of black enamel paint (Rustoleum) was sprayed directly on the surface to generate a high-contrast speckle pattern. The enamel paint contains solvents that may affect the properties of the tissue, but due to the small coating thickness and short period of testing time (t⩽30 min), the potential influence was ignored. In addition, the speckles were nearly dry on contact with the tissue, which further limited potential effects from the solvent. After deposition, there was minimal evidence of running or diffusion of the paint.

The specimens were placed within a dedicated load frame and subjected to uniaxial monotonic tensile loads under displacement control actuation. Compression grips were used to clamp both ends of the specimens and the average gage length between grips was 75 mm. Displacement increments of 1.27 mm (0.05 in.) were applied to the aorta sections until reaching a maximum axial deformation of 25 mm. The reaction load resulting from elongation was determined using a precision load cell with full-scale range of 22.5 N. An image of the specimen’s surface was recorded prior to loading and at each step of axial displacement. The tension experiments for the arterial specimens were comprised of loading the specimen to 25 nominal elongation and unloading 12 times consecutively. The 13th and 14th cycles consisted of loading to 30 and to 40 elongation and unloading, respectively. The first 12 cycles were conducted to condition the samples and achieve a steady state mechanical response. Speckle images were documented during the 1st, 4th, 8th, 12th, 13th, and 14th cycles at every 2 elongation along with the corresponding axial load. A typical experiment comprised of the 14 load-unload cycles with one specimen required about 30 min.

An image size of 25×18 mm was utilized and digitized into a sample of 1280×960 pixels with 256 gray levels. A comparison of the digitized speckle distributions captured at each load step was conducted to determine the full-field displacement by means of DIC. The strain distribution was then calculated from the displacement distribution according to Eq. (4). A typical stress/strain response for an arterial section at the 14th cycle is shown in Fig. 4(a). The elastic moduli for the axial sections of arterial wall were determined using the tangent method from the stress and strain relations in each cycle and using the average strain over the entire window of evaluation. The average primary (E₁) and secondary (E₂) elastic moduli for the arterial specimens was 192 and 912 kPa, respectively. An example of the axial strain distribution over the surface of the tissue at 40 elongation is shown in Fig. 4(b). Note the distinct patterns in strain distribution, as evident from the gray-scale variation. The coefficient of variation in displacement and corresponding strain was approximately 0.16 regardless of the degree of elongation. The elastic modulus reported for specimens comprised of intimal and medial layers of the descending aorta is³² 248 kPa. Mechanical properties of arteries are dominated by the medial layer as it is comprised of the highest percentage of collagen fibers. The specimens examined in this study were comprised of the intima and a portion of the media; the average elastic modulus after conditioning was 192 kPa. Based on the difference in ratio of intimal to medial components, it is reasonable that the elastic modulus was less than that reported in the literature. Note, however, that the axial stress was calculated using actual cross-sectional area, based on the assumption that volume is preserved.

Figure 4

Typical response of a bovine artery subjected to uniaxial tension: (a) stress/strain response of a bovine arterial section during cycle 14 and (b) axial strain distribution in the arterial wall.

With proper surface preparation of the substrate, DIC provides a useful method for evaluating the constitutive behavior of soft tissues over a large range in strain. Based on results from the analysis of bovine aorta, DIC can be used to examine the constitutive behavior of soft tissues and the variability that may result from structural variations or pathogenesis. It was also found that DIC is relatively insensitive to whether the surface is wet or dry.²⁵ The full-field characterization enabled by DIC may provide new knowledge regarding the structure/property relationship for arterial tissues.

4.2.

Mechanical Behavior of Bovine Hoof Horn

Lameness of dairy cows resulting from hoof and leg ailments is a major problem in the dairy industry. A preliminary study of the effects of moisture on the elastic properties of bovine hoof horn was recently conducted to evaluate the potential effects of evolving dairy conditions on hoof injuries.^{33 34} Hooves were obtained from mature Holstein and Black Angus cows within 12 h of sacrifice. Thin uniform slices of approximately 0.5 mm were excised incrementally from the sole toward the dermal-epidermal junction of both the lateral and medial claws of the front and rear hooves. Conventional dogbone tensile specimens were then stamped from the individual slices, as shown in Fig. 5(a). In this preliminary study, the elastic properties were evaluated according to two levels of hydration, which were achieved according to specific storage conditions. All specimens were stored dry at 4 °C for a period no less than 1 week; the “dry” samples were stored in air while the “wet” samples were maintained in a bath of deionized water. Although the moisture content was not quantified, the specimens were considered dehydrated and fully saturated (100 relative humidity). According to the natural pigmentation of hoof horn, the raw specimens were dark gray or black. After removing the specimens from storage, a thin uniform coat of matte white enamel paint was sprayed over the gauge section and was followed by a peripheral coating of black enamel paint. The combination resulted in a high-contrast speckle distribution suitable for DIC. All samples were tested immediately after removal from storage.

Figure 5

Preparing tensile specimens from the sole of cow hooves: (a) excising thin sections from hoof claw and (b) geometry of typical tension specimens stamped from the serial sections.

The hoof horn specimens were subjected to uniaxial tension under displacement control using a universal testing machine (Instron Dynamite Universal Testing System, Model 8841, Canton, Massachusetts). Displacement increments of 1.27 mm (0.05 in.)/min were applied to the specimens until reaching a maximum axial load of 20 N or failure of the specimen. The reaction load resulting from elongation was determined using a precision load cell with full-scale range of 22.5 N. The image capturing system was initialized by a trigger signal originating from the universal test machine at the onset of loading. Sequential images with spatial resolution of 640×480 pixels and 256 gray scales were captured from an evaluation window of 8.5×6.4 mm every 2 s until reaching an axial load of 20 N or failure of the specimen. The displacement/strain distribution can be deduced using DIC in one of two ways. One is to analyze each image in real time while the images are being documented. In this manner, the amount of pixels being analyzed is limited by the computer’s processing speed and the time interval between sequential images. The other approach is to store the sequential images onto the hard drive while performing the tension experiments and then to process them consecutively after the experiment. The second approach was adopted as the strain can be deduced with increased precision using a larger subset area for establishing correlation.

A typical stress/strain response for wet and dry specimens is shown in Fig. 6(a). The elastic moduli for the specimens in both dry and wet conditions was determined from the elastic response for strains less than 0.1 using the tangent method. The average elastic modulus of the “dry” hoof horn was 2870±460 MPa, which is significantly larger than that of the “wet” horn (E=102±34 MPa). Results from this preliminary study indicate that environmental conditions at the dairy farm may be important to the stiffness of cow hoof horn and reductions in stiffness with increasing moisture may raise the potential for hoof injuries. Poisson’s ratios for the hoof horn specimens in the two conditions (wet and dry) are shown in Fig. 6(b). Note that the largest variation in Poisson’s ratio occurs at the onset of axial loading and is attributed to the small transverse component of displacement. Similar to the stiffness, Poisson’s ratio for the hoof horn is a function of the moisture content and the implications of this change on hoof health remain to be addressed. A typical full-field description of the axial strain distribution over the entire window of evaluation (8.5×6.4 mm) for a tensile specimen is shown in Fig. 6(c). Similar to the strain distribution for the bovine artery, there are distinct variations in the axial strain indicating an inhomogeneous distribution in the mechanical properties. The variation in strain is evident from the gray-scale map and suggests that there are abnormalities of the tissue. The average coefficient of variation in the axial strain distribution for all specimens was 0.18. Using the full-field characterization available from DIC it is possible to establish stochastic descriptions for the elastic properties of cow hoof horn and to define the mechanical behavior as a function of the hoof anatomy. These valuable details will be examined in future studies.

Figure 6

Mechanical behavior of the hoof horn established using DIC: (a) typical stress/strain response for wet and dry hoof horn, (b) Poisson’s ratio for the wet and dry specimens form (a), and (c) strain distribution of wet cow hoof horn under uniaxial tension

4.3.

Loosening in Cemented Total Hip Replacements

Loosening of the femoral component in total hip replacements (THR) is one of the most common sources of hip replacement failure. The influence of bone surface preparation on the mechanics of loosening in cemented joint replacements was studied. The investigation relies on the use of DIC to identify changes in relative displacement between the bone, cement, and prosthesis of model THR systems.^{35 36}

All experiments were conducted with fresh bovine femurs obtained within 12 h of sacrifice. The middiaphyseal portion of the femur was sectioned to obtain a specimen 66 to 76 mm in length. Sections with relatively uniform canal diameter (d<30 mm) and a thick wall (t>7 mm) were then selected for further preparation and evaluation. The sectioned bone was milled to 63.5 mm long, and the canal diameter was increased to 31.8 mm, using a specific tool chosen to impart a desired surface profile and corresponding roughness. A model commercially pure titanium (CPTi) prosthesis is then inserted within the prepared femoral section and cemented in place using polymethylmethacrylate (PMMA) bone cement (DePuy Endurance® PMMA bone cement, Warsaw, Indiana) and advanced cementing techniques (water lavage and cement pressurization). The CPTI prosthesis has a 25.4-mm diameter and knurled surface to simulate the rough texture of a real prosthesis. Dimensions of the femoral canal and prosthesis result in a cement mantle thickness of approximately 3.2 mm. Although the prosthesis morphology is slightly different than that of commercial devices, it provides an adequate model with which to reconstruct the cemented total hip replacement.

The cemented joints were further modified to enable an examination of the interface displacement. An evaluation window was milled in two opposing sides of the specimen [Fig. 7(a)] to reveal the cement/bone and cement/prosthesis interfaces [Fig. 7(b)]. Flaws may be introduced within the cement during machining of the evaluation windows. Although these flaws may reduce the fatigue life they have been ignored since the importance of bone surface preparation on loosening is quantified in terms of the relative difference in fatigue life rather than the absolute number of cycles to failure. The windows serve as a venue to acquire images that encompass the interfaces and enable quantification of the relative displacement between the prosthesis, cement, and bone that result from axial loading of the prosthesis. Based on the specimen configuration there are four independent interface pairs that can be evaluated. The windows provide a direct view of the interfaces and represent a 2-D surface that supports application of DIC. Surface preparation is then required to minimize the glare of the titanium surface and to establish a high-contrast speckle field. For convenience of application, a white correction fluid was used to coat the surface, followed by a peripheral spray of black enamel paint to deposit the desired black dots.

Figure 7

Model femoral hip replacement with titanium prosthesis (b=bone, c=cement, p=titanium prosthesis): (a) model implant with evaluation window revealing the prosthesis/cement and cement bone interfaces and (b) the interfaces of the specimen in (a) with specific bone surface preparation.

After implanting the prosthesis and surface preparation the specimens were subjected to a monotonic axial load in seven equal increments ranging from +100 to −667 N. Loads were applied through the titanium prosthesis in a dedicated load frame while the specimen was supported vertically beneath the femoral wall. Note that a concentric axial load applied to the model joint while eccentric loading is more likely in vivo. The load configuration was simplified to enable a clear distinction of the effects from surface preparation on the loosening process. When the significance of bone surface preparation on loosening using axial loading is accomplished, future efforts will evaluate loosening under a more complete load distribution.

In all experiments, the digital camera was placed normal to the specimen surface at a distance capturing the bone/bone cement and bone cement/titanium prosthesis interfaces, as shown in Fig. 8. A digital image was captured at each increment of loading, compared with an image captured at no load, and then processed to determine the displacement distribution. The images comprised an area of 19×14 mm and were digitized into 2048×1536 pixels with 256 gray scales; the corresponding resolution in displacement was 10 μm/pixel. A typical full-field displacement distribution across the interface of a model joint replacement is shown in Fig. 9(a). The interfaces are clearly evident from the gray-scale map. The displacement distribution along a specific horizontal line extending across the two interfaces is shown in Fig. 9(b). The process was followed by fatigue loading of the model cemented joints with an axial load of +100 N to −667 N in increments of 1 million cycles. At the completion of 1 million cycles, the specimen was placed on the dedicated load frame, loaded axially, and the full-field displacement distribution was characterized using DIC.

Figure 8

Experimental setup for the evaluation of loosening in THR.

Figure 9

Displacement distribution of a model implant system after fatigue loading (b=bone, c=cement, p=titanium prosthesis): (a) gray-scale map representing the axial displacement (parallel to the prosthesis) and (b) axial displacement distribution across the interfaces.

The incremental fatigue and evaluation process enabled quantitative distinction of the initiation and progression of loosening in the cemented joints. More importantly, the mechanisms of loosening and the dependence on bone surface preparation and texture can be established from the full-field characterization of displacement. Preliminary attempts at characterizing the mechanisms of loosening in the cemented joints were conducting³⁵ using ESPI. However, the limited range of displacement measurement and difficulties with the surface preparation required for ESPI in the moist conditions proved DIC to be a much more suitable optical method. Therefore, in this application, DIC provides extremely valuable information over a range of deformation that cannot be acquired using other optical techniques. With further investigation it may be possible to identify surgical bone surface preparations that maximize the longevity of cemented total joint replacements.