Introduction

Total hip surgery is a procedure performed for hip pain relief using either cemented or uncemented approaches; each of these approaches has advantages and disadvantages. The most important disadvantage of the cemented method is osteolysis associated with cement disease [1, 2], whereas stress shielding and high cost are the disadvantages of the uncemented method [3, 4]. It is very difficult to decide which method is more successful based on the literature because of discrepancies among the published results. Some studies point out that the cemented method leads to higher revision rates in the first 10 years [5, 6]; that is exactly opposite of other studies [7–9]. In a similar way, some studies reported that total hip surgeries using the uncemented method have both a higher survival [3] and revision rates [9].

The studies show that whether the cemented or uncemented methods is superior one another since many factors such as stem design, patient age, bone quality, surgeon experiences, etc. [5] affected as seen in Table 1. However, the most important factor is the stem design, since recently the stem design have been developed due to clinical experiences [10, 11]. Also, the studies reported that the uncemented method should be preferred in young patients and the cemented method should be applied in older patients [10, 12–15]. There were no differences between the cemented and uncemented techniques used in clinical studies in terms of the survivorship [16, 17].

Table 1 Comparisons of the clinical studies in the literature about the cemented and uncemented methods
Full size table

In this study, both methods, cemented and uncemented, were evaluated in terms of mechanical aspects using finite element analysis (FEA). In the FEA processes, the most important variables, the cement usage, and stem materials were considered. The equivalent stress and micro-motion between the stem and femur were compared to evaluate which one is the best for the hip surgery in view of different mechanical aspects.

Materials and methods

Model development

The femur was modeled in 3D using computerized tomography (CT) images obtained from a male patient (aged 63, Body-Mass Index 26 Kg/m2) taken from a Toshiba Aquilion CT scanner in the Department of Radiology, School of Medicine at Kocaeli University. CT images consist of parallel layers having a pixel size of 0.774 × 0.774 mm at the lateral position and a voxel resolution of 473 × 473 × 235. 1841-layer shootings were carried out to develop the model, recorded in the Digital Imaging and Communications in Medicine (DICOM) format, and transferred to the MIMICS 17 (Materialise, Leuven, Belgium) 3D image-processing software. The surface errors such as spikes, intersections etc. of the femur models were corrected using Geomagic Studio 10 software (Raindrop Inc., USA). After these corrections, the 3D smooth solid model was developed and imported into SolidWorks program (Dassault Systems SolidWorks Corp., USA) in IGES format.

The cemented and uncemented stems were modeled in SolidWorks using design parameters taken from the TIPSAN Catalog (TIPSAN Co. Inc., İzmir/Turkey). The cement thickness was created as 2 mm to obtain better results in the modeling processes [18–20]. The created femur head was operated according to surgery procedure and combined with the cement and stem. The combined femur and stem models were transferred into the ANSYS Workbench (ANSYS Inc. Canonsburg, PA) in STEP format.

Material properties mesh and contact assignments

It was assumed that the material properties of all models were selected as linear, elastic and isotropic. The most commonly used two different materials, Ti6Al4V and CoCrMo, were defined and imported their properties into the ANSYS Workbench Software as seen in Table 2 [21, 22]. The modulus of elasticity for the femur was described according to the density that was calculated from the CT images in MIMICS [22].

Table 2 The material properties of the models [21, 22]
Full size table

Mesh convergence was tested by refining the element size from 6 to 3 at 1 mm interval on the femur, 4 to 1 at 0.5 mm interval on the stem and 2 to 0.5 at 0.25 mm interval on the cement models. The most suitable element sizes for the optimum results were determined as 4, 1.5 and 0.75 mm for the whole femur, stem, and cement, respectively. Solid187 tetrahedron element was used in the whole finite element model, in which several mesh sizes along with additional refinements were defined at contact regions to get convergences. The number of the elements are changing between 400,000 and 800,000 and the nodes are changing from 600,000 to 950,000 in the FEA models, as seen in Fig. 1. Ten different models were named with their abbreviations of the methods, stem material names and contact definitions as seen in Table 3. The cement and bone interface was assumed to be bonded because of perfect connection in the cemented models [18, 21, 23–25]. The interface was also assumed frictional contact with the frictional coefficient of 1.0 [26]. The cement-stem interfaces were presumed to be either bonded [23] or frictional contact type [26]. The stem-bone contact types for the uncemented models were defined as either frictional [27, 28] or bonded contact [29, 30].

Fig. 1
figure 1

The femur-stem models along with boundary conditions

Full size image
Table 3 The model names considering the materials and contact types
Full size table

Boundary conditions

The boundary conditions were defined for the models as seen in Fig. 1. The stem-bone models were subjected to static loads obtained from literature in accordance with the value reported for a person walking at a normal speed. The coordinate system for the femur was defined based on the definition by Bergmann et al. [31]. Considering the body weight, the maximum forces resulting from walking were applied to the femoral head. The muscle forces were defined as presented by Duda et al. [32] (Table 4). The distal end of the femur was constrained in three directions considering the contact surface of the knee joint.

Table 4 The forces applied to the femur considering 700 N weight person (see Fig. 1)
Full size table

Results

Three different criteria were selected to evaluate the results obtained from the FEA. First, the von Mises stresses on the stem, femur contact surfaces and cement were calculated to assess the strength. Second, the relative micro-motions were investigated between the stem and femur to evaluate the formation of the osseointegration in the uncemented models [33]. Finally, the axial normal strain distributions of the femur were calculated to examine the stress shielding [29, 30, 34, 35].

The effects of the stem material

The stem material affected the stress distribution and micro-motion on the femur contact surfaces depending on the contact conditions. As can be seen Table 5, the Ti6Al4V alloy stem reduced the peak von Mises contact stress (PVMCS) on the femur surface among the uncemented models when comparing the CoCrMo Alloy stem (UcTiFri versus UcCoFri). However, the Ti6Al4V alloy stem increased the PVMCS of the femur in the cemented models with the frictional contact when compared to CoCrMo Alloy stems (CTiFriFri versus CCoFriFri and CTiFriBo versus CCoFriBo). Besides, the Ti6Al4V stem decreased the PVMCS of the femur in the models with bonded contact. Ti6Al4V stem usage caused more PVMS on the cement than CoCrMo stem usage in frictional contact conditions (CTiFriFri versus CCoFriFri and CTiFriBo versus CCoFriBo). However, Ti6Al4V stem decreased the peak von Mises stress (PVMS) on the cement in bonded contact (CTiBoBo versus CCoBoBo). The PVMS on the Ti6Al4V stems has lower than the CoCrMo stems as expected. The maximum PVMS of the stem was obtained lower than the yield stress of the materials as shown in Table 2. Therefore, the mechanical failures of stems were not expected according to these conditions. Osseointegration was available for all the uncemented models considering the peak micro-motion according to the study of Jasty et al. (lower than 40 microns) [33].

Table 5 The results of the cemented and uncemented models
Full size table

The effects of the contact conditions

The PVMCS of the femur surfaces were affected due to the contact types between stem and femur. As illustrated in Table 5, the PVMCS in the uncemented models obtained approximately four times the bonded contacts comparing the frictional contacts. However, the PVMCS of femur surfaces decreased in the bonded contacts versus frictional one in the cemented models. The PVMS on the stem increased in the bonded contacts. The contact type between stem and cement or cement and bone also affected the PVMS of cement. The PVMS of cement with bonded contacts was calculated lower than with frictional contacts. The lowest PVMS of cement was obtained in the CTiBoBo model. The micro-motion between bone and stem was calculated higher in frictional contact as expected.

The effects of the cement

As shown in Table 5, the cement usage with Ti6Al4V stem increased the PVMCS on the femur surfaces in the frictional contacts (UcTiFri versus CTiFriFri). When using the contact was bonded, cement usage with Ti6Al4V decreased the PVMCS on the femur surfaces (UcTiBo versus CTiFriBo and CTiBoBo). The cement usage with CoCrMo stem decreased the PMVCS of the femur in both frictional and bonded contacts (UcCoFri versus CCoFriFri, UcCoBo versus CCoFriBo and CCoBoBo). Finally, the cement usage caused to decrease the PVMCS of femur surfaces.

The cement usage with Ti6Al4V did not affect the PVMS of the stems in the frictional contacts (UcTiFri versus CTiFriFri and CTiFriBo). When the contact was bonded, the cement usage with Ti6Al4V decreased the PVMS of stems (UcTiBo versus UcTiBoBo). The cement usage with CoCrMo stem decreased the PVMS of stems in both frictional and bonded contacts as seen in Table 5 (UcCoFri versus CCoFriFri and CCoFriBo, UcCoBo versus CCoBoBo). Therefore, the cement usage causes to decrease the PVMS on the stem.

The effects of the stress shielding

The axial normal strains were calculated at ten different designated points on the models as seen Fig. 2. The highest compressive strain was found on the intact (non-surgical) femur. The stem insertion caused to decrease strain values especially at point five and six. The cement usage increased the strains on the femur that decreased the stress shielding. Less stress shielding was obtained in the Ti6Al4V stem comparing the CoCrMo stem. The different contact types are also causing the stress shielding emerge in all implanted models. Therefore, the frictional contact caused to decrease the strain values more than bonded contacts.

Fig. 2
figure 2

The calculated axial normal strains for selected points

Full size image

Discussion

The selection of proper method, cemented or uncemented, considering the stress shielding and aseptic loosening issues is still unclear regarding clinical studies since it depends on surgeon experiences. The clinical treatment observations after hip surgeries can be evaluated numerically using engineering techniques such as FEA since the stress shielding and aseptic loosening are the most important issues for failing the hip surgeries [5, 29]. As mentioned above, there is no biomechanical study comparing those methods considering failure issues. For these reasons, both techniques were evaluated using two different materials (Ti6Al4V and CoCrMo) in biomechanical viewpoints.

According to the Wolff’s law [34], the meaning of stress shielding is the reduction of the stresses on the femur due to changing load transfer after the stem insertion. That is; the removal of stresses from the bone refers to the weak bone density. However, when the stem is implanted, the load is transferred through the stem rather than the femur, that results in the bone weakening and stress shielding. In addition, the stem design parameters and material types play important roles for the stress shielding problem [30, 34]. The stress shielding values are increasing when using the CoCrMo stem material as seen in Fig. 2. The stress transferring from the stem to the femur is much easier when using the Ti6Al4V having higher ductility as given Table 2. Therefore, the stem made out of stiffer material causes higher stress shielding as seen in Table 5. These results obtained in the study also confirmed the studies given in the literature [35]. Lower stress shielding was also obtained in cemented fixation technique rather than uncemented one, also confirmed by [29], since both cement and cancellous of the femur have similar material properties. Using the cement in the fixation process increases the strain values at the proximal part of the femur. The contact types also affect the load transfer between the femur and stem as obtained in the FEA process as illustrated in Fig. 2. The frictional contact is defined to represent the healing (post operation) process and bonded contact is defined to simulate the healed femur (osseointegrated femur). Therefore, the load transfer is more suitable in the healed femur to decelerate the stress shielding occurrence.

The strength of the cement is very important since the crack growth in the cement plays an important role for the generation of aseptic loosening [16]. Using the stiffer stem material reduces the stress occurrence in the cement as reported similarly in the literature [36, 37]. In the FEA process, the higher stress values leading crack growth were obtained in the cement when using Ti6Al4V material along with frictional contact. Therefore, using of CoCrMo material gives better results during the healing process, but the Ti6Al4V material is more convenient for healed femur in terms of nonformation of cement cracks.

The contact types of the cement-stem and cement-bone couples were described differently in the literature. The cement-bone interfaces were mostly described as either bonded contact type due to perfect bonding in the healing (post operation) process [18, 21, 23–25] or frictional contact by Ramaniraka et al. [26]. The frictional contact and bonded contact types were used by [18, 21, 23, 25, 26], respectively, between the cement-stem models. Therefore, all contact types were considered in our study, as given in Table 3. While the stress (PVMCS) values on the femur and (PVMS) on the cement decreased using the bonded contact but increased using the frictional contact. In the uncemented models, however, the stress occurrences are the other way around. Based on these obtained results, uncemented fixation method has lower survival rate comparing the cemented fixation technique which was also reported by Morshed et al. [38]. But, cemented fixation technique has riskier due to cement crack formation, using the frictional contact type that causes the revisions of the hip fixation also reported by Abdulkarim et al. study [16]. On the other hand, the femur contact stresses (PVMCS) evaluated in this study (Table 5) were not considered by the literature [21, 24, 29]. The higher PVMCS (Table 5) values on the contact surfaces cause femur contact surface failure that generates the small gaps and then results in the aseptic loosening.

After total hip fixation, the failures of the stem do not or rarely occur since the calculated PVMS values obtained changing from 150 to 565 MPa lower than the yield stress of the stem materials. This result was also confirmed by researchers [39, 40]. The micro-motion values between the stem and femur should not exceed the 40 microns for ideal osseointegration formation. If the micro-motion values exceed the 40 microns and lower than 150 microns, the fibrous tissues are taken place that weakens the bonding of the bone-stem couple. If the values exceed the 150 microns, then undesired aseptic loosening occurs [33]. Based on the results obtained in the study, the ideal osseointegration formation is available in uncemented fixation technique. Therefore, when using the uncemented fixation method, the risk of the aseptic loosening is rather low during the healing process.

In this study, some assumptions were used that may be limited the results. First, the values of loading conditions were taken into account during the only regular walking in the FEA modeling process. Second, the design parameters of the selected stem model taken from the TIPSAN Co. cataloque were also considered in both cemented and uncemented fixation modeling. Third, the bonding interfaces of the cement-femur, cement-stem, and femur-stem couples were assumed to be bonded perfectly but the contact ratio is variable depends on patients, surgeon experiences, stem, and cement qualities. Finally, the material properties of the bone models were defined as isotropic but the bone materials are usually accepted as anisotropic [41]. The isotropic material properties for bone can be acceptable because of small differences between the isotropic and anisotropic material properties of bone [42]. The definition of the bone material properties as isotropic in the modeling was also depended on the MIMICS Software. However, different boundary conditions, stem design parameters, and material properties may be used in the modeling process that may change the results slightly.

Conclusion

In this study, two different methods, cemented and uncemented for the stem fixation in hip surgeries using two different materials for the stem were compared and analyzed with the use of computerized FEA simulations. Different boundary conditions were also considered in the simulations. Based on the study, the following results can be concluded.

  • For the stem fixation in hip surgery, cement material usage reduces the stress shielding and aseptic loosening problems.

  • The stiffer material (CoCrMo) used for the stem reduces the stresses on the cement during the healing process but causes the cement cracks formation in healed femur. Therefore, the stiffer material used for the stem have shorter survivorship.

  • The stiffer stem material (CoCrMo) used for the stem causes higher stress shielding in the femur, therefore, the Ti6Al4V material selection for the stem is more convenient.

  • It was observed that the definition of contact types in the modeling process plays an important role in the results. Therefore, the contact types should be selected considering the modeling conditions.

The cement material usage in hip surgery is more suitable considering the stress shielding and aseptic loosening problems.