ELSEVIER 1350-4533(95)00067-4

Med. Eng. Phys. Vol. 18, No. 5, 396404, pp. 1996 Copyright 0 1996 Elsevier Science Ltd for IPEMB

Printed in Great Britain. All rights reserved 1350-4533/96 $15.00 + 0.00

Dynamic fracture characteristics of the osteochondral junction undergoing shear deformation

N. D. Broom, A. Oloyede, R. Flachsmann and M. Hows

Biomechanics Laboratory, Department of Mechanical Engineering, University of Auckland, New Zealand

Received 12 December 1994, accepted 10 April 1995

ABSTRACT This paper presents a biomechanical study of the dynamic fracture response of the osteochondral regions of both

immature and mature cartilage-bone laminates induced through shear loading. An instrumented impact machine,

providing both mechanical and reabtime macrophotographic fracture data, was used to load the osteochondral region

in shear by means of a direct compressive force applied to the cartilage layer only, and in a direction parallel to the

plane of the osteochondral junction. This force was applied at a known velocity of 2 500 mm/s. Our results show

that under these conditions of shear loading the dynamic mode II fracture toughness of the osteochondral region of

the mature tissue is approximately 1.5 times that of its immature counterpart, and that the derived dynamic shear

stiffness of the immature tissue is about 4 times that of the mature. The structural studies demonstrated that the

mature tissue delaminates within a we&deJined tidemark region whilst the immature fractures through the subchond-

ral bone into which fingers of articular cartilage penetrate. We suggest that this pseudobrittleness of the immature

tissue could explain why there is an increased susceptibility to osteochondral failure in the younger human joint.

Finally, we note that the strengtktotoughness relationship, in which stiJfm engineering materials are known to exhibit

lowerfracture toughness, also applies to the cartilage-bone system. Copyright 0 1996 Elsevier Science Ltd for IPEMB.

Keywords Osteochondral region, impact loading, pure shear deformation, maturity-dependent failure mode, toughness

Med. Eng. Phys., 1996, Vol. 18, 396-404, July

INTRODUCTION

The cartilage-bone tissues within the articulating joint form a durable laminate structure capable of sustaining a wide range of loading conditions. A critical region of this complex tissue system is undoubtedly the osteochondral junction where the more compliant cartilage is structurally inte- grated with the much stiffer subchondral bone. Any interface linking a material of a given com- pliance to another material with a very different compliance will represent a region where large and potentially damaging shear forces may be developed when the composite system is mechan- ically loaded.

Although the primary forces acting on the joint surface during normal function will be mainly compressive, secondary shear forces are likely to be generated in the osteochondral region. as a direct consequence of the above-mentioned modulus mismatch effect. Our laboratory has

Correspondence to: Dr Neil Broom, Biomechanics Laboratory, Department of Mechanical Engineering, University of Auckland, Private Bag 92019, Auckland, New Zealand.

shown in some earlier experiments that direct impulsive loading of the articular surface of carti- lage-bone samples will induce delamination of the cartilage layer in the vicinity of the osteochon- dral junction’ and this is most probably a conse- quence of very high levels of shear forces being generated.

In the more relevant area of whole joint behav- iour where the pattern of forces acting on the joint tissues are likely to be highly complex, stud- ies have shown that when the joint is either twisted or dislocated, convex articular surfaces can be exposed to large lateral forces which may lead to the development of various shear-induced lesions in the osteochondral region*“. Tissue maturity is widely believed to influence the joint’s suscepti- bility to osteochondral failure. Previous investi- gators have noted that the plane of cartilage sep- aration in the elderly adult tends to occur along the calcified-uncalcified cartilage interface (i.e. the tidemark), leaving the osteochondral junction undisturbed5. By contrast, the adolescentjoint has no calcified layer, therefore shear forces may be transmitted deep into the osteochondral region with consequent subchondral fractureGp7.

During the adolescent growth spurt there is an increased risk of subchondral fracture which may in some instances progress towards subarticular post-traumatic necrosis or osteochondritis disse- canPlo. Edwards and Bentley”, and Desai et ~1.‘~ have postulated that osteochondritis dissecans pat- ellae is a consequence of repeated partial dislo- cation of the patella in which shear forces prod- uce osteochondral fragments still attached at one side. Similarly, Livesley and Milligan’” concluded that the abnormal pattern of loading associated with the recurrent subluxation of the patella, which would involve shear forces, is the most likely cause of osteochondritis dissecans.

In view of this probable link between shear forces, related deformation, and ~YZ viva failure of the osteochondral region it would seem that the mechanical parameters of shear strength and shear toughness could provide a fundamental biomechanical characterization of the junction region of this tissue system. In particular, the values of these parameters for both immature and mature tissues at high loading rates are likely to have relevance to the clinical situation.

In a recent paperI we reported the response of the osteochondral region to unconstrained shear loading at impact velocities. Whilst this study yielded both quantitative and qualitative insights into the behaviour of this region under the imposed mechanical conditions it did not allow the evaluation of purely in-plane shear fracture parameters which could uniquely characterize the osteochondral region. This is because the uncon- strained loading method used in that earlier study resulted in complex fracture of the osteochondral region which combined both failure modes I and II as defined in classical fracture mechanics (see Figure I). Specifically, the lack of constraint on the deformation of the cartilage layer in the direction perpendicular to the direct loading axis encour- aged a substantial tensile mode I failure in which a proportion of the applied energy was dissipated through the out-of-plane and intra-matrix tearing of the overlying cartilage layer’“. Consequently, in order to obtain mode II fracture parameters such as the dynamic fracture toughness, shear failure strength, etc., which can quantifjr the degree of susceptibility to fracture of this important region, we must impose conditions which will lead to in-

OPENING

MODE I

plane sliding failure only, and which confine crack growth within the osteochondral region.

In this present study we have disallowed mode I fracture in the osteochondral region by using a modified indenter to impact the cartilage layer in transverse compression, such that the experimen- tal data obtained can be used to determine the sliding dynamic fracture parameters for osteo- chondral failure in both immature and mature tis- sue systems.

Behaviour of the articular cartilage-bone laminate under impact loading

At high rates of deformation close to and includ- ing unpact the stress-strain response of cartilage has been shown to be effectively independent of strain-rate, thus indicating an elastic response”. Consequently, we conclude that it will respond in a similar manner when impacted as part of the cartilage-bone laminate system. Also, bone is well known to respond in a linear elastic manner under a high strain-rate/low strain’“.‘7 regime, and we assume that it will respond in the same manner under the present loading conditions. Because of these individual responses, the carti- lage-bone laminate will be anaiysed as an elastic system, and the interspacing osteochondral region analysed using linear elastic fracture mech- anics (LEFM). Furthermore, we argue that the imposed constraints and loading conditions are such that:

1. The major portion of the energy supplied to the system by impacting the cartilage layer is used up in fracturing the osteochondral region, since this anchoring region will mostly resist the applied load; and

2. Crack propagation will occur as a wholly mode II fracture process induced through the lateral sliding of the deforming cartilage layer.

Pertinent fracture mechanics of the osteochondral region

The fracture mechanics parameters relevant to this present investigation are summarized as fol- lows.

(a) The dynamic energy release rate or dynamic

SHEARING

MODE II

397

Impact shear fracture of asteachondral region: N. Lt. Broom et al.

fracture toughness in mode II shear, C;,,‘“, which is evaluated from

(1)

is the strain energy release rate as an initiated crack is propagated to failure. This change in energy a Wis evaluated as the area under the load- displacement curve which is obtained directly from our instrumented impact testing system. B is the thickness of the specimen and this has been chosen as half the width of each specimen (see Tissue preparation), and aa is the measured change in crack length.

In using equation (1) we have assumed that under impact the osteochondral region, whether in the immature or mature tissue, responds in a linearly isotropic, elastic manner during fracture. We note, however, that this will result in an approximation of the values of toughness obtained since the actual cartilage-bone system is multilayered, anisotropic, and inhomogeneous.

(b) The dynamic critical stress intensity factor in mode II sheaP, Z&,, is given by

KIId = YT 6, (2)

where Y is the geometry correction factor which is used to modify the infinite plate solution r&a in order for it to apply to a finite geometry. Although of finite geometry, Y has been assigned a value of 1 since we are restricting the present analysis to one of plane stress. We also note that equation (2) represents the relationship for calcu- lating the fracture toughness of strictly linearly elastic and isotropic materials. The assumed value for Y, and the equation’s application to a laminate system, albeit with components exhibiting near- elastic properties at impact, mean that only rela- tive values of KIId are obtained, and these must therefore be treated with caution. However, we argue that these approximations should not lessen their usefulness as indicators of relative fracture toughness in the present analysis. T is the shear stress required to propagate a crack of length a to failure, and ‘TT is the standard constant. r was calculated from the impacting force divided by the area of the osteochondral plane (i.e. the area of the articular surface).

Using equations (1) and (2), the dynamic shear modulus18 of the osteochondral region which we term n/r, can be evaluated as

The parameters GiId, KIId and @ will be evaluated from the results of the impact loading exper- iments which are described in the following sec- tion. However, it should be noted that because K IId cannot be determined absolutely, the expression for A& in equation (3) must be viewed as an approximate adaptation of fracture mech- anics at this stage.

EXPERIMENTAL METHOD

Tissue preparation

Macroscopically normal patellae were removed from freshly slain 2-3-year-old bovine animals (immature tissue), and approximately 5-7-year- old bovine animals (mature tissue). The tissue was stored at -20°C. Samples of articular cartilage-on- bone of dimensions in the plane of the articular surface approximately 4 x 8 mm and of variable depth were rough sawn from the medial side of each patella. To obtain rectangularly shaped, smooth-sided specimens, the sawn samples were ground on successive grades of Carborundum paper, from 80 to 600 grit, under running water. The samples were then placed in 0.15 M saline solution for at least 30 min at room temperature. During hydration the exposed edges of the articu- lar cartilage layer tended to swell laterally, obscur- ing the otherwise distinct osteochondral region. Therefore just prior to testing the samples were reground with 600 grit paper in 0.15 M saline sol- ution. The final preparatory step involved using a fine water-soluble marker pen to draw a pattern of parallel lines on the side face of the specimen where the fracture process was to be photo- graphed during impact. This pattern of lines was used to delineate the osteochondral region of the specimen, making the crack length and defor- mation pattern easily identifiable on the macro- photographs. The appearance of the inked grid pattern also provided a simple method of con- ducting a preliminary check on a sample’s maturity. The ink trace, on traversing the imma- ture junction region, exhibited an irregular stain- ing appearance created by the regions of soft, more easily stained cartilage which interdigitated with the microscopically roughened texture of bone which, in turn, stained far less uniformly (see for example Figure 3b). By contrast, the much more planar junction between the uncalcified and calcified cartilage, i.e. the tidemark, characteriz- ing the mature tissue samples yielded a more crisp appearance of the grid lines in the osteochondral region (see for example Figure 5b). This prelimi- nary method of classification was subsequently checked by histology and found to be extremely reliable (see Microscopy).

Impact testing procedure

The impact loading arrangement is shown sche- matically in Figure 2. It consisted of a vertical-faced indenter incorporating an adjustable horizontal upper extension which effectively constrained the delaminating cartilage layer so as to prevent any upward movement. The indenter was fixed to the free end of a rigid pendulum arm that could be released from a predetermined height so as to impact the cartilage-bone test specimen in trans- verse compression at the lowest point of swing, and at a previously calibrated velocity of 2500 mm/s (i.e. a nominal strain rate of 250 s-l).

The precise positioning of the cartilage-bone sample with respect to the pendulum swing was

398

Impnr-1 shear fracture of osteorhondrul ygion: IV. Il.

UPPER CONSTRAINING FACE OF INDENTER

Broom et al.

Figure 2 Schematic of impact loading arrangement usrd to load the ostrochondral I-egicm in pure shear

achieved by using a specialized mount with 2 degrees of freedom, which allowed complete alignment of the cartilage-bone interface with respect to the lower edge of the indenter face. This lower edge was positioned approximately 0.2 mm above the delineated osteochondral junction to ensure direct transverse loading of the cartilage only. The upper constraining extension was adjusted so that it was parallel to and just touching the articular surface of the cartilage.

To measure the impact load response a Kistler piezoelectric stress washer (type 9021, 35 000 N) was incorporated into the indenter assembly. The output signal was fed into a Kistler charge ampli- fier (type 5007). The load signal from the charge amplifier was then passed though a 10 kHz filter.

A visual record of the test specimen was obtained just prior to impact and during impact using a 35 mm camera fitted with a macro lens and mounted at right angles to the axis of loading so as to focus on the side face of the sample that had been marked with the parallel lines. The specimen was illuminated using a stroboscope set in the single flash mode with a flash duration of 1.7 ps. To trigger the stroboscope a knife edge extension of the pendulum arm passed though an adjustable optical switch (type RS304-560) with a time delay of less than 10 ps. The flash signal was detected using a fotonic sensor (MT1 model KD- 100). This fotonic sensor utilized an optical probe (type KD 285 H) with a frequency response of 150 kHz, and produced a transient voltage peak at the instant the stroboscope flashed, thus indicating the precise moment the photo was taken. The impact experiment was conducted with the cam- era shutter held open, and in a darkened room.

The load signal and the output from the fotonic sensor were simultaneously recorded on a dual- trace digital storage oscilloscope, thus providing an exact correlation between instantaneous frac- ture load and position of the advancing crack. The oscilloscope traces were triggered by the initial load rise from the force transducer. From a knowledge of the indenter velocity, both the force-time and fotonic voltage-time traces were

converted to force and voltageAisplacement curves for subsequent quantitative analysis of the fracture process.

Test sample numbers

A full mechanical analysis was carried out on 24 separate test specimens prepared from nine immature patellae and on 17 test specimens pre- pared from eight mature patellae.

Microscopy

Structural examination of the impacted tissues was conducted on six of the immature and six of the mature specimens in order to confirm the maturity of the samples and to establish the exact location of their failure planes in the osteochond- ral region. The specimens were fixed in 10% buff- ered formalin CPC, decalcified in calex, followed by paraffin wax mounting. Sections 6 pm in thick- ness were cut as osteochondral cross-sections and stained with toluidine blue.

RESULTS AND DISCUSSION

General

Under the imposed conditions of constrained shear deformation substantially different responses were observed between the immature and mature tissue specimens, and these are evalu- ated in the following.

Fig-w-es >5 show mechanical and macro-photo- graphic data obtained from the impact of imma- ture and mature cartilage-bone samples, respect- ively. These were approximately representative of all the tissue responses obtained in the present study. The mean nominal shear stresses, which were evaluated as the ratio of the maximum trans- versely applied compressive load (determined from the force-displacement curves) to the total area of the osteochondral plane of a given speci- men, are 6.4 (SE. = 0.2) MPa and 6.4 (S.E. = 0.4) MPa for the immature and mature tissues, respect-

399

Impact shear fracture of osteochondral region: N. D. Broom et al. . (4 -FOTONIC PULSE

FORCE

-!z 0 2.5 5.0

DISPLACEMENT (MM)

Figure 3 (a) Force-displacement curve obtained from impacting an immature tissue test sample. Position of pulse on fotonic trace indi- cates the instant at which the macro-photograph in (c) was taken. (b) Sample just prior to impact loading. (c) Sample as in (b) undergoing shear fracture. Arrowhead markers in (b) and (c) indi- cate exact location of a ‘landmark’ feature in subchondral bone on grid line 4. The horizontal displacement of this feature relative to the chondral portion of the line in (c) shows that actual shear frac- ture has occurred at least up to and including this grid line. Note minimal shear distortion of cartilage layer in (c). One scale division = 0.5 mm

iveiy. These similar values of applied shear stress indicate that the observed fracture mechanisms in both the immature and mature tissues which are described below are a consequence of comparable loading conditions.

Within the framework of classical fracture

(4

! FOTOHIC PULSE

2.5 5.0

DISPLACRMEEIT (MM)

Figure 4 (a) Force-displacement curve obtained from impacting a mature tissue test sample. Pulse on fotonic trace indicates the instant at which the macro-photograph in (c) was taken. (b) Sample just prior to impact loading. (c) Sample as in (b) undergoing shear frac- ture. Note extensive shear distortion of cartilage in vicinity of tidem- ark. One scale division = 0.5 mm

mechanics the force-displacement curves shown in Figures 3u, 4a and 5a can be interpreted as fol- lows: crack initiation will occur during the stage of loading leading up to the point of maximum load, whilst the crack propagation stage will be represented by the unloading portion, i.e. from the maximum load down to zero (see Figures 3u, 4a, 54.

Accurate positioning of the horizontal con- straining extension of the indenter is clearly effec- tive in preventing any upward motion of the

lmpntl shearfraclure of’osteochondrul region: LV. I). Broom et al.

300

r

_ FOTONIC PULSE

0 2.5 5.0 7.5

DISPLACEHENT (MM)

Figure 5 (a) Force-displacement curve obtained from impacting a mature tissue test sample. Pulse on fotonic trace indicates the exact moment at which the macro-photograph in (c) was taken. (b) Sam- ple just prior to impact loading. (c) Sample as in (b) undergoing shear fracture. Note extensive shear deformation of cartilage in vicin- ity of tidemark. One scale division = 0.5 mm

delaminated section of cartilage (see Figures 3c, 4c, 5c), thus showing that a combination of the imposed constraint on the cartilage surface and transverse compression yields the desired shear deformation of the osteochondral region. Con- versely, Figure 6 illustrates how a tensile opening mode is introduced into the delamination process when the gap between the constraining face and the articular surface is excessive.

Immature tissue response. Figure 36 and c provides a clear delineation of the extent of the shear frac- ture ahead of the indenter face loading the imma- ture tissue. A careful inspection of the ‘landmarks’ on the lower portions of the vertical grid lines

Figure 6 Example of mixed modes 1 and II type fracture of imma- ture tissue specimen resulting from too large a clearance between articular surface and horizontal constraining surface of indenter. One scale division = 0.5 mm

delineating the subchondral bone, and their hori- zontal positions relative to the corresponding upper parts of the grid lines before (Figure 36) and during delamination (Figure 3~) indicates that continuity across the osteochondral region is sev- ered up to and including grid line 4. This indi- cated that the mode II crack has advanced to a position somewhere between lines 4 and 5, i.e. approximately 3.5 mm ahead of the indenting face, thus providing an instantaneous crack length. This represents delamination of approxi- mately 70% of the osteochondral plane, or a con- dition of catastrophic/fast fracture. Correlation of the force-displacement (Figure 3a) and macro- photographic data indicates that this substantial length of mode II crack generated ahead of the indenter corresponds to a stress close to the maximum value attained during the test and that minimal additional work was required to com- plete the shear delamination process. As is evident from the small amount of angular reorientation of the vertical grid lines at this stage of crack growth, the delamination was achieved with only minimal shear distortion of the cartilage layer.

Mature tissue response. The macro-photographic data obtained from the mature tissue tests (Figures 46, c and 56, c) reveal substantial differences in the mechanism of crack formation compared to that in the immature tissue. Specifically we observe a much greater resistance to crack propa- gation. Examination of the grid pattern in Figure 4c shows that the subchondral end of line 4 ident- ified with an arrow corresponds almost exactly with the position of the indenting face (obscured by a fragment of delaminated cartilage but indi- cated by vertical broken line). Although the curved appearance of this line reveals a high level of shear distortion of the cartilage matrix in the osteochondral region, its structural continuity with the subchondral bone is maintained thus indicating that the shear crack at this instant had not propagated ahead of the indenting face. In Fipw 5c continuity of the severely distorted line

401

Impact shear fracture of osteochondral region: N. L). Broom et al.

No. 3 (identified with an arrow) is maintained across the osteochondral junction. Although it is difficult to see clearly whether or not there is a crack already formed between lines 1 and 3, the presence of a fragment of cartilage ahead of the indenter in this picture indicates that some delamination has occurred in this initial stage of loading. However, the indenter itself has already advanced almost to line no. 5, thus indicating that the propagating crack lags considerably behind the advancing indenter, i.e. extensive defor- mation of the articular cartilage has occurred in the vicinity of the osteochondral region without actual shear failure taking place.

Furthermore, the force-displacement curves in Figures 4a and 5a show very slow rates of unloading from the peak load. In fact, Figure 4a shows that the load continued to increase slightly following the initial peak, and this is inconsistent with the predictions of fracture mechanics. It may well be that we observe this fundamental departure from theory because of the highly complex structure of the osteochondral region. Possibly, the nature of interaction of the applied stress with the structure encourages multiple initiation of cracks which then join to form a crack of sufficient length to cause final failure. However, this can only be speculation at this stage.

The above observations suggest that there is a higher level of resistance to crack formation and growth in the mature osteochondral region than in the immature. Also, the severe grid line distor- tion in the mature osteochondral region relative to the bulk of the cartilage layer demonstrates that it is primarily the osteochondral region that resists the advancing indenter under the imposed con- ditions of loading. This confirms our earlier assumption (1) in the subsection Behaviour of the articular cartilage-bone laminate under impact loading.

Microscopic examination of fracture planes

Light microscopic examination of the sections obtained from the immature tissue samples indi- cated that the osteochondral shear fracture propa- gated along an irregular plane within the region of predominantly subchondral bone containing interpenetrating fingers of articular cartilage (Figure 7). There was a complete absence of any tidemark, i.e. the distinct junction between the uncalcified and calcified cartilage which charac- terizes the fully mature joint system.

By contrast, the mature tissue samples fractured either along the relatively smooth plane of a well- developed tidemark or in the uncalcified articular cartilage itself in the vicinity of the tidemark. Fig- ure 8 illustrates both tidemark delamination and chondral failure in the same mature tissue sample. The shear fracture has propagated mostly along the tidemark but the final stages of delamination have been completed via chondral failure. These observations of the planes of fracture in the imma- ture and mature tissue systems are entirely consist- ent with our earlier study in which we showed that

delamination occurred via mixed mode fracture in contrast to the present mode II mechanism.

DYNAMIC FRACTURE MECHANICS PARAMETERS FOR THE OSTEOCHONDRAL REGION

Fracture toughness

Using equation (1)) the average values of fracture toughness C;,, for the immature and mature tissue osteochondral junctions were evaluated as 0.036 (S.E. = 0.002) MN/m and 0.058 (SE. = 0.002) MN/m, respectively. These values indicate that disparities between samples within each group are not statistically significant. It should be noted that in using equation (1)) the area under each force- displacement curve a W was taken as the area of the region described from the first instance of loading to the displacement corresponding to the transient pulse on the fotonic trace. This is because the crack length, which was obtained from noting the extent of grid-line discontinuity in the macro-photographs, and substituted for au in the calculation corresponded to the energy released only up to the displacement at which the fotonic sensor pulse was recorded. These values of fracture toughness show that the mature tissue is approximately 1.5 times tougher than the imma- ture.

Dynamic stress intensity factor, K,,

This fracture parameter which characterizes the stress state at the tip of a growing crack was calcu- lated using equation (2) and with Y= 1. The mean values of this parameter for both the immature and mature samples are 0.72 (SE. = 0.03) MN/m3’* and 0.44 (S.E. = 0.03) MN/m”“, respectively. Again, we see that there is no signifi- cant disparity between individual samples within each of the two groups. These values of KIId reveal that the stress field responsible for crack propa- gation in the immature tissue is about 1.6 times greater than that for the mature tissue. It should also be noted here that there can be no absolute correlation made between KIId and GIId since the first is sensitive to geometry. Also, the value of KIld cannot be determined absolutely using equation (2) because the correct value for Yin the present study is unknown.

Dynamic shear modulus of rigidity, ik&

This parameter was derived from equation (3)) and values of 15.8 (SE.= 1.4) MPa and 3.7 (S.E.= 0.5) MPa were obtained for the immature and mature tissues, respectively. These values indicate that the dynamic shear stiffness of the immature tissue’s osteochondral junction is approximately 4 times that of the mature.

Comparison of mode II only to mixed mode fracture parameters

For the mixed mode fracture reported in our pre- vious paper14, the governing dynamic strain

402

Figure 7 Typical irregular appearance of-osteochondral fracture surface in immature tissue. Note fingers of cartilage (A<:) which anchor- into the subrhondral hone (B)

’ 0.4mm ’ J

Figure 8 Typically smooth appearance of predominantly tidemark (I‘) frzacture surface in mature tissue system. Final stageb of fracture have occurrrd in thr unralcified cartilage (AC) rather than along thr tidcmarh srpamting the uncalrified from the calcified (CC) cxtilagr

energy release rates and hence fracture toughness values G’ were 0.008 MN/m and 0.016 MN/m for the immature and mature tissues, respectively. When compared with our present values of dynamic sliding fracture toughness Gld, we note that Cd for the immature tissue is an order of mag- nitude less than GILd, while Gild = 4C$ for the mature osteochondral region. These relation- ships, we believe, suggest that it is much more dif- ficult to initiate and subsequently propagate a pure shear crack within the osteochondral regions of either of these tissue systems than it is to do so under mixed mode conditions.

CONCLUSIONS

In this study we have characterized the in vitro dynamic fracture responses in pure shear of the

immature and mature osteochondral junctions found in the bovine animal joint, and have com- pared these responses with those obtained under mixed mode fracture.

The typical mode II fracture response observed in the immature tissue system is that of cata- strophic fracture, whereas the mature tissue sys- tem possesses a greater ability to resist the growth of an initiated crack; i.e. it is characterized by a relatively higher dynamic fracture toughness C;,,.

We therefore argue that the toughness or ability to resist crack growth exhibited by the mature tis- sue system is a consequence of the energy absorb- ing capability of its more resilient tidemark region relative to that of the immature tissue in which this structural feature is absent. This greater toughness of the mature tissue is also borne out by the higher amplitude (i.e. A?,,<,) of its crack tip

403

Impact shear fracture of osteochondral region: N. D. Broom et al.

stress field, which is required to propagate a crack in its osteochondral region. By contrast, the cata- strophic fracture observed in the immature tissue can be attributed to the stiffer, but less tough bone encountered by the advancing crack as it propagates osteochondrally.

Finally, the system of testing we have utilized in these studies is one of pure shear, and is therefore a considerable over simplification of the mechan- ical conditions that are likely to prevail in the functioningjoint. Also, it is acknowledged that the assumption of linear, isotropic elasticity can only be a crude approximation for the complex carti- lage-bone system. However, our experiments do indicate very clearly that, given the same idealized conditions of testing of the immature and mature tissue systems, there is a substantial difference in their respective values of mode II fracture tough- ness. And because there is clinical evidence sug- gesting that shear forces may play an important role in osteochondral breakdown, we believe the present study represents an important contri- bution to the biomechanics ofjoint tissue failure.

ACKNOWLEDGEMENTS

This project was funded by a project grant from the Health Research Council of New Zealand.

REFERENCES

1. Silyn-Roberts H, Broom ND. Fracture behaviour of carti- lage+m-bone in response to repeated impact loading. Connect Tissue Res 1990; 24: 143-56.

2. Johnson-Nurse C, Dandy DJ. Fracture-separation of articular cartilage in the adult knee. J Bone Joint Surg 1985; 67B: 42-3.

3. Tomatsu T, Imai N, Takeuchi N, Takahashi K, Kimura N. Experimentally produced fractures of articular carti- lage and bone. J Bone Joint Surg 1992; 74B: 457-62.

4. Imai N, Tomatsu T, Okamoto H, Nakamura Y. Clinical and roentgenological studies on malalignment disorders

of the patell*femoral joint. Part III: Lesions of the patellar cartilage and subchondral bone associated with fz$lo-femoral malalignment. JJup OrthOp Assoc 1989; 63:

5. Meachim G, Bentley G. Horizontal splitting in patellar articular cartilage. Arthritis Rheum 1978; 21: 669-74.

6. Matthewson MH, Dandy DJ. Osteochondral fractures of the lateral femoral condyle. J Bone Joint Surg 1978; 60B: 199-202.

7. Rosenberg NJ. Osteochondral fractures of the lateral femoral condyle. J. Bone Joint Surg 1964; 46A: 1013-26.

8. Chan D, Aldridge MJ, Maffulli N, Davies AM. Chronic stress injuries of the elbow in young gymnasts. Br JRudiol 1991; 64: 1113-8.

9. De Smet AA, Fisher DR, Graf BK, Lange RH. Osteochon- dritis of the knee: value of MR imaging in determining lesion stability and the presence of articular cartilage defects. Am JRoent 1990; 155: 549-53.

10. Hughston JC, Hergenroeder PT, Courtenay BG. Osteo- chondritis dissecans of the femoral condyles. J Bone Joint

Surg 1984; 66.4~ 1340-8. 11. Edwards DH, Bentley G. Osteochondritis patellae. J Bone

Joint Surg 1977; 59B: 58-63. 12. Desai SS, Pate1 MR, Michelli LJ, Silver JW, Lidge RT.

Osteochondritis dissecans of the patella. J Bone Joint Surg 1987; 69B: 320-5.

13. Livesley PJ, Milligan GF. Osteochondritis dissecans patel- lae. Is there a genetic predisposition? Znt Ortho$19?2; 16: 126-9.

14. Flachsmann R, Broom ND, Oloyede A. A biomechanical investigation of unconstrained shear failure of the osteo- chondral region under impact loading. Clin Biomech 1995; 10: 15&65.

15. Oloyede A, Flachsmann R, Broom ND. The dramatic influence of loading velocity on the compressive response of articular cartilage. Connect Tissue Res 1992; 27: 21 l-24.

16. iizkaya N, Nordin M. Fundamentals of Biomechanics -

Equilibn’um, Motion and Deformation. Van Nostrand Rein- hold, New York, 1991: 343-7.

17. Viano DC. Biomechanics of bone and tissue. A review of material properties and failure characteristics. Proceed- ings - SA,?. SAE, Warrendale PA, 1986; P186: 33-63.

18. Ewalds HL, Wanhill RJH. Fracture Mechanics. Edward Arnold, London, 1984; 28-116.

404