Tensile Characteristics of Ten Commercial Acrylic Bone Cements

E. J. Harper, W. Boneld
IRC in Biomedical Materials, Queen Mary and Westeld College, University of London, Mile End Road, London, E1 4NS, United Kingdom

Received 23 July 1999; revised 27 March 2000; accepted 27 March 2000

Abstract: The mechanical properties of acrylic bone cement, used in orthopedic surgery, are very inuential in determining successful long-term stability of a prosthesis. A large number of commercial formulations are available, differing in chemical composition and physical properties of both powder and monomer constituents. In this study, the static and dynamic tensile characteristics of a number of the most commonly used bone cements (Palacos R, Simplex P, CMW 1 & 3, Sulx-60, Zimmer Dough), along with some newer formulations (Endurance, Duracem 3, Osteobond and Boneloc), have been investigated under the same testing regimes. Testing was performed in air at room temperature. Signicant differences in both static and fatigue properties were found between the various bone cements. Tensile tests revealed that Palacos R, Sulx-60, and Simplex P had the highest values of ultimate tensile strength, closely followed by CMW 3, while Zimmer Dough cement had the lowest strength. Fatigue testing was performed under stress control, using sinusoidal loading in tension tension, with an upper stress level of 22MPa. The two outstanding cements when tested in these cyclic conditions were Simplex P and Palacos R, with the highest values of Weibull median cycles to failure. Boneloc bone cement demonstrated the lowest cycles to failure. While the testing regimes were not designed to replicate exact conditions experienced by the bone cement mantle in vivo, there was a correlation between these results and clinical outcome.
2000 John Wiley & Sons, Inc. J Biomed Mater Res (Appl Biomater) 53: 605 616, 2000

Keywords: ture

acrylic bone cement; cemented arthroplasties; mechanical testing; fatigue; frac-

INTRODUCTION Bone cement is used as a grout to x implants in place during joint replacement surgery. A polymer powder based upon poly(methylmethacrylate) (PMMA) or a related co-polymer is mixed in surgery with a monomer, usually methylmethacrylate (MMA). Chemical and physical processes occur simultaneously, resulting in a doughy mass, which is inserted into the prepared cavity. The material sets, stabilizing the implant approximately 1520 min after the initial mixing. In the body, bone cement is subjected to a repetitive loading pattern.1 Although bone cement is reasonably strong in compression, it is a relatively brittle material, making it susceptible to fracture as a result of tensile loads. It is not surprising, therefore, that bone cement has been implicated as one of the factors that causes aseptic loosening.2-4 The Swedish National Hip Registry found aseptic loosening to be the most common reason for revising a hip replacement, producing 73.2% of all revisions recorded between 1979 1996.5

Correspondence to: Dr. E. J. Harper, IRC in Biomedical Materials, Queen Mary and Westeld College, University of London, Mile End Road, London, E1 4NS, UK (e-mail: e.j.harper@qmw.ac.uk) Contract grant sponsor: EPSRC 2000 John Wiley & Sons, Inc.

Following the introduction of bone cement by Sir John Charnley in the 1950s, there have been numerous investigations into its fatigue properties, which have been comprehensively reviewed by Krause and Mathis6 (from 1974 1987) and Lewis7 (from 19871997). These reviews demonstrate the large variety of fatigue protocols that have been followed and the limited number of cements studied using any given testing technique. This background makes it almost impossible to compare results from different investigations. In this current investigation, the same fatigue testing method has been used to assess the tensile fatigue behavior of several of the commercial cements in clinical practice to provide an independent assessment of their relative fatigue behavior. Mechanical tests can be chosen to suit particular cement types, which do not reveal important characteristics of the material. This is especially important for the introduction of new cement formulations, since the ISO standard for orthopedic bone cement8 does not at present include a comprehensive mechanical testing program. Due to the cyclic stresses bone cement is subjected to in vivo, fatigue properties of bone cement are an important factor in the long-term survival of a cemented hip replacement. The experimental procedure employed in this study was not designed to be the same as the physiological environment of bone cement. However, there
605

606

HARPER AND BONFIELD

was a direct correlation between the results of the fatigue testing and clinical data reported in the Swedish National Hip Register.5

MATERIALS AND METHODS

Commercial Cements

The bone cements used in this study included a combination of established commercial cements and some newer cements introduced into the market more recently. The cements tested, and their manufacturers, were Palacos R (E. Merck, Darmstadt, Germany), Surgical Simplex P (Howmedica International Ltd., London, UK), CMW Types 1 & 3 and Endurance (Depuy Ltd., Blackpool, UK), Zimmer dough type and Osteobond copolymer cement (Zimmer, Warsaw, IN), Sulx-60 (Sulzer, Winterthur, Switzerland), Duracem 3 (SulzerMedica, Sulzer Orthopaedics Ltd., Baar, Switzerland) and Boneloc (Polymers Reconstructive A/S., Farum, Denmark). Boneloc has now been withdrawn from the market following some early loosening of joint prostheses. Although each cement is mainly based upon a PMMA homopolymer or MMA copolymer with an MMA monomer, all cements have a distinctive formulation leading to different handling and resultant mechanical properties. A list of the components of each polymer powder and liquid is given in Table I; this information is taken directly from the manufacturers information included in the cement packaging.
Preparation of Cements

specimen extension. The cross-head speed employed was 5 mm/min, and the maximum force recorded was used to obtain the ultimate tensile strength, ult. A value for secant modulus, Esec, was taken at 10 MPa and strain at failure, f, was also calculated. At least ve specimens were tested, and the mean and standard deviation is reported for each cement. The fatigue tests were performed on an MTS 810 electrohydraulic testing machine and cycled continuously in load control until failure. The cyclic stress employed was sinusoidal at a frequency of 2 Hz. Ten specimens of each cement were tested in tensiontension with a lower stress of 0.3 MPa and an upper stress of 22 MPa. All the mechanical testing was conducted in air at room temperature. A Weibull model was used to represent the fatigue data graphically. The fatigue lives were sorted in ascending order and each datum point was assigned a median rank value, P, obtained from a statistical table.9 This value was used to calculate a Weibull number, W, for each fatigue life using the following equation: W log(1/(1-P )) The Weibull number was plotted against cycles to failure, using a logarithmic scale, to produce a straight line representation of each set of fatigue data. A value for Weibull median was calculated for each cement type at 50% failure probability.
Scanning Electron Microscopy

All the cements were prepared according to the manufacturers instructions at 23C; this procedure was very similar for each cement. The polymer powder was placed in a clean glass beaker, and the monomer was added and stirred using a spatula until the powder was fully wetted. The time this took varied with each cement, but was always less than 40 s. The mixture was subsequently either transferred to the syringe body of a cement gun and injected into a PTFE mould at approximately dough time, usually about 2 min, or manually inserted into the mould, according to the manufacturers instructions. The lled moulds were pressurized to 1.4 MPa and held there until the cement had hardened, approximately 15 min. The exception to this procedure was Palacos R, which was precooled to 4C prior to mixing. A maximum of eight half-sized ISO 527 multipurpose test specimens were produced from each 40 g powder sachet. The samples were 75 mm in length, 5 mm in width, approximately 3.5 mm in thickness, with a gauge length of 25 mm. A schematic diagram of the specimen design is shown in Figure 1. Each sample was measured and stored for at least one week at 37C in dry conditions prior to testing.
Mechanical Testing

Electron microscopy was carried out using a JEOL scanning electron microscope. Both tensile and fatigue fracture surfaces were examined after the application of a gold coating using an accelerating voltage of 10 kV. Micrographs of details of interest were taken at a range of magnications.

RESULTS Table II shows the results of the tensile testing, giving mean values for ultimate strength, secant modulus, and strain at failure. Numbers shown in brackets are the standard deviations. The cements are ranked according to the mean value obtained for the ultimate strength. Statistical analysis was used to assess differences between the maximum strengths of the cements using a student t-test and the results are displayed in Table III. The results from the fatigue testing, Weibull median, and range of cycles to failure are displayed in Table IV. These cements have been ranked according to the value obtained for the median cycles to failure. Again statistical analysis was performed, applying the MannWhitney Utest,10 and, in this case, results are shown in Table V. The distributions of cycles to failure for each cement are shown on the Weibull plot in Figure 2 (a). Figures 2(b) and (c) show the same results on a different scale to enable differences in the distributions to be observed more easily. Scanning electron micrographs of the various fatigue fracture surfaces are

Tensile testing was conducted on an Instron Testing Machine, Model No. 6025 with a clip-on extensometer to measure

TENSILE CHARACTERISTICS OF ACRYLIC BONE CEMENTS TABLE I. Compositions of Commercial Bone Cements Tested

607

Key: Polymers: PMMA-poly(methylmethacrylate) P(MMA/MA)-methylmethacrylate/methacrylate copolymer P(MMA/sty)-methylmethacrylate/styrene copolymer P(MMA/BMA)-methylmethacrylate/butylmethacrylate copolymer Initiator: BPO-benzoyl peroxide Monomers: MMA-methylmethacrylate BMA-butylmethacrylate DCMA-n-decyl methacrylate IBMA-isobornyl methacrylate Accelerators: DMT-N,N-dimethyl-p-toluidine DMPE-N,N-dimethyl-amino-phenethanol DHPT-Dihydroxyl-propyl-p-toluidine

608

HARPER AND BONFIELD

greater than 2 mm was very low. The majority of pores, if present, were less than 1 mm in diameter.
Comparison of the Strengths of the Cements

Figure 1. Dimensions of tensile and fatigue test specimens.

shown in Figures 310. Only micrographs exhibiting features of interest are included.

DISCUSSION The handling characteristics of each cement varied considerably, thus requiring differing amounts of mixing to fully wet the powder. Since all mixing in this particular study was performed in air, this methodology resulted in a varying degree of porosity among the cements. This observation was made via a visual examination of fracture surfaces; no detailed porosity measurements were conducted. It was decided that it was more clinically relevant to include all samples that were possible to test, in the tensile and fatigue testing, rather than use any specimen rejection criteria, as has been used in previous studies.11 Therefore, if failure occurred in the gauge length, the value was taken to be a valid result. In practice, the number of specimens resulting in pores in the fracture surface

The values from the static tensile testing highlighted the large distribution of properties exhibited by commercial cements on the market and in clinical use. There was a wide range of tensile strength values: Palacos R, Simplex P, and Sulx-60 gave the highest values of strength of approximately 50 MPa, CMW 3 gave a value of 44.7 MPa, while CMW 1, Boneloc, Osteobond, and Endurance produced strengths of approximately 40 MPa. The value for Zimmer dough type was the lowest at 31.7 MPa. The Palacos R, Simplex P, and Sulx-60 cements were signicantly higher in strength compared to the other cements tested with the exception of CMW3. There was no statistical difference between the values obtained for CMW 1, Boneloc, Osteobond, and Endurance. The Young modulus results ranged from 2.26 GPa for Boneloc to 3.53 GPa for CMW1. Values for the strain to failure varied from 1.36% for CMW3 to 2.48% for Boneloc. The differences among fatigue results for the different cements were much larger than those found with the static tensile results. The highest Weibull median fatigue cycles to failure obtained for Simplex P and Palacos R were considerably higher than found for Zimmer dough type and Boneloc. What is also important is the distribution of fatigue lives. A narrow spread of fatigue lives is better than a wider scatter of data, because it indicates greater predictability in vivo. There is some correlation between the static and fatigue

TABLE II. Static Tensile Properties of Commercial Bone Cements

TABLE III. Statistical Differences Between Ultimate Tensile Strengths of Commercial Cements Obtained Using the Student t-test TENSILE CHARACTERISTICS OF ACRYLIC BONE CEMENTS

609

610

HARPER AND BONFIELD TABLE IV. Fatigue Results of Commercial Bone Cements

strengths, but the ranking of static strength does not exactly follow that of the fatigue lives.
Correlation of Results to Clinical Data

Factors Affecting Strength

The method of fatigue testing employed in this investigation is not designed to be a replica of the exact conditions experienced by a bone cement mantle surrounding a hip replacement. The test does, however, assess the tensile fatigue properties of the various cements tested and, since bone cement is weaker in tension compared to compression, it is more likely to fail due to the tensile component of the cyclic loads when subjected to them in vivo. Therefore, some correlation between the fatigue results and clinical data is to be expected. The majority of investigations studying the loosening of hip replacements do not compare the different cements tested. An exception to this is the Swedish Hip Register.5 In the 1998 review of 148,359 primary hip replacements, the authors reported, Lowest risks are associated with Pallacos Gentamycin, plain Pallacos and Simplex. CMW has slightly worse result with the highest risks associated with Sulx. This order of success is the same as that obtained from the fatigue test carried out in this investigation, which is encouraging. It should also be noted that the cement with the signicantly lowest fatigue properties is Boneloc, a material that was withdrawn from the market due to its high incidence of loosening.

The reasons for the differences obtained in mechanical properties can be attributed to variations in both composition of polymer and monomer, particle size, morphology, and molecular weight of powder, strength of polymer bead-matrix interface, and powder-to-liquid ratios. Cracks grow within bone cement intergranularly (i.e., in the newly formed polymeric phase), transgranularly (i.e., in the preformed polymeric bead), and along the bead-matrix interface.12,13 Therefore, both the powder and liquid components are important. Another important inuence upon the mechanical properties is the method of sterilization. The sterilization technique used for the majority of the cements is gamma irradiation, with the exception of Palacos R, which is ethylene oxide sterilized, and Sulx-60 and Duracem 3, which are sterilized via formaldehyde tablets. It has been shown in previous work that gamma irradiation causes a large decrease in the strength of a bone cement due to loss in molecular weight,14,15 whereas the use of ethylene oxide and formaldehyde does not affect mechanical strength. Because there are many factors inuencing cement strength, it is not easy to interpret the varying strengths of the cements. The main constituent of Simplex P is a P(MMA/styrene) copolymer and this composition results in one of the highest values of both tensile and fatigue strength. Osteobond copolymer cement is also composed of a P(MMA/styrene) copolymer, but, although

TABLE V. Statistical Differences between Fatigue Properties of Commercial Cements Obtained Using the MannWhitney U-Test

612

HARPER AND BONFIELD

Figure 2. Weibull distributions of fatigue cycles to failure after being cycled 0.322 MPa for (a) all cements tested, (b) common commercial cements, and (c) newer commercial cements: () Simplex; () Duracem; () CMW1; () Boneloc; () Palacos; () CMW3; () Endurance; ()Osteobond; () Sulx-60; () Zimmer dough.

this also has a relatively high fatigue resistance, it is not signicantly different from Duracem 3, CMW 3, and Sulx-60, which are chiey composed of PMMA. One major factor is the ratio of PMMA to styrene in the copolymer, and this information was not supplied. Palacos R also has outstanding strength, with a relatively narrow distribution of fatigue lives. The high strength may be attributed both to the inuence of the P(MMA/MA) copolymer and to the method

of sterilisation, i.e., via ethylene oxide, which does not have a detrimental effect upon mechanical properties. The distribution of fatigue life is inuenced most by the presence of aws in the material and the resistance to propagation of these aws. Palacos R possessed a similar porosity on the fracture surface to Simplex P, as assessed visually, suggesting that its better distribution of fatigue life was due to it being a tougher cement compared to most of the other ce-

TENSILE CHARACTERISTICS OF ACRYLIC BONE CEMENTS

613

Figure 2. (continued)

ments tested.16 It is also important to note that, with the exception of Boneloc, Palacos displayed the highest strain to failure. Another factor to consider is the value obtained for cement modulus. Crowninshield et al.17 showed, using a 3-dimensional nite element analysis model, that a lower cement modulus results in lower stresses experienced by the cement in vivo. However, the value of modulus does have to be viewed in the context of both tensile and fatigue strengths. In general, the cements with the highest values for modulus possessed the higher values of tensile and fatigue strengths.
Comparison to Literature Studies

Relatively few investigations are reported in the literature comparing a wide range of mechanical properties of com-

mercial bone cements using similar test protocols. A study by Kusy, 1978,18 compared the tensile strengths of CMW, Palacos, with and without gentamicin, Sulx-6 and Simplex P after 1 month and after conditioning for 10 month in distilled water. The highest strength after 1 month was found for the Sulx-6 and Palacos, the lowest for the Simplex P. After aging in distilled water, all the values of strength were reduced, with the highest values for the CMW and Palacos with gentamicin. There have been several investigations into the mechanical properties of Simplex P compared to Zimmer dough and LVC, the most commonly used cements in the USA. Weber and Bargar, 1983,19 reported no statistical differences between the tensile strengths of these three cements after 14 days cure, but the Zimmer cements displayed lower exural strengths. When tested in exure, Simplex P gave the highest result. In agreement with this result, was a report by Davies et al., 1987,20 who reported no differences in

Figure 3. Fatigue fracture surface of Palacos R bone cement.

Figure 4. Fatigue fracture surface of Simplex P bone cement.

614

HARPER AND BONFIELD

Figure 5. Fatigue fracture surface of CMW 1 bone cement.

Figure 7. Fatigue fracture surface of Osteobond co-polymer bone cement.

strength when the cements were tested in tension. However, after testing in tension-compression fatigue, under stress control, Simplex P possessed superior fatigue properties. Data published by Gates et al., 1984,21 and Krause et al., 1988,22 showed that Zimmer cements had inferior fatigue properties as compared to Simplex. In a study comparing the fatigue characteristics of Palacos R and Simplex P by Davies et al., 1989,23 no signicant difference was obtained. A more recent study by KindtLarsen et al., 1995,16 investigated a range of commercial cements including CMW-1, Palacos R, Simplex P, Zimmer dough and LVC, and Boneloc. When mixed in an open bowl, Palacos R displayed the highest value of tensile strength, followed by Simplex P; Zimmer dough had the lowest strength. In exure, Simplex P had the highest value followed by Palacos R, again Zimmer dough gave the lowest result. The fracture toughness data showed Palacos R with highest value, the other cements giving similar values, with the exception of Boneloc, which gave a value almost 50% lower. Fatigue properties of Simplex P in comparison to Boneloc were also reported. When tested in strain control, Boneloc displayed the highest cycles to failure, whereas in stress control, Simplex P was superior. In view of the history of Boneloc in

clinical use in the body,24 fatigue testing under stress control appears to be a better indicator of a prediction of clinical success. In summary, Simplex P and Palacos R generally possessed the highest strengths for both static and fatigue properties, and Zimmer dough displayed the lowest values, in agreement with results reported in this article for a wider range of cements. The review of previous studies highlights the problems of comparing data from different investigations, which use varying methods and only a limited number of bone cement formulations.
Scanning Electron Microscopy

The fatigue fracture surfaces examined via SEM revealed, in general, relatively little plastic deformation occurred upon fracture resulting in at fracture surfaces. There was also, in the majority of cases, evidence of the barium sulphate or zirconium dioxide particles, added as a radiopaque ller, having been pulled out of the surface upon failure. Failure was often initiated from an internal pore. Figure 3 shows a micrograph of a Palacos R cement fracture surface, revealing zirconia particle pullout and relatively smooth regions of bead fracture. The zirconia particles

Figure 6. Fatigue fracture surface of CMW 1 bone cement at a higher magnication.

Figure 8. Fatigue fracture surface of Sulx-60 bone cement.

TENSILE CHARACTERISTICS OF ACRYLIC BONE CEMENTS

615

tended to agglomerate and were much larger in size compared to the barium sulphate particles used in other cements. In the region around the pore that initiated failure, the surface was rougher with more zirconia pull-out, indicating slow fracture. Further from the pore, towards the edge of the sample the surface was smoother, indicating fast fracture. Figure 4 is a typical micrograph obtained for Simplex P cement, in which it was possible to observe fracture around preformed beads surrounded by the barium sulphate. This observation indicated that fracture had occurred through both the preformed bead and newly formed inter-bead matrix, as described by previous researchers.12,13 Barium sulphate did not show evidence of agglomeration, unlike the zirconia in the Palacos R. The cements CMW 1, CMW 3, and Endurance appeared similar when viewed via the SEM. Figure 5 is a typical example of the fracture surfaces, which were similar to Figure 4. The CMW 3 cement surface was slightly rougher in appearance, indicating the fracture was less brittle than the CMW 1 and Endurance cements. Figure 6 shows a higher magnication micrograph of the barium sulphate particles. The particles appear to be situated in pores within the PMMA matrix, with no evidence of any bond between the two opposing surfaces. Zimmer dough cement fracture surfaces were similar to both the CMW and Simplex P. There was, however, evidence of considerable porosity on the fracture surfaces. Osteobond cement contained less pores compared to the Zimmer cement and less barium sulphate pullout. Figure 7 shows a typical fracture surface revealing a rougher surface compared to Figure 4. Sulx-60 cement fracture surfaces were different in appearance from the other cements; there was little ceramic particle pull-out and the surface gave evidence of less brittle failure. This observation was supported by the strain-to-failure results shown in Table II. The radiopacifying agent was zirconia, and particles are shown in Figure 8. Duracem 3 cement gave similar results to Sulx-60, not surprising since the compositions are very similar. Boneloc cement fracture surfaces were different from the majority of other cement fracture surfaces. The surfaces showed evidence of a more ductile failure and there were regions where there appeared to be separate layers of

Figure 10. Fatigue fracture surface of Boneloc bone cement showing zirconia pull-out.

material, as shown in Figure 9. There was very little evidence of particle pull-out, but circular holes were observed in some areas, from which, it was assumed, zirconia particles had been pulled out. Some spherical particles remained in these holes, as shown in Figure 10.

CONCLUSIONS This study revealed that tensile and, in particular, fatigue tests highlighted large differences in the strengths of the commercial bone cements investigated. The cements that perform best clinically gave the highest results in this study. In view of this result, it appears important to test all experimental bone cements in prescribed cyclic testing regimes in order to evaluate their fatigue performance prior to use in surgery. This initial study used only one fatigue test condition for sterilized cements prepared by hand mixing. However, the value of the present study is in demonstrating independently the wide range of fatigue performance in commercial bone cements. To evaluate the cements fully, a comprehensive study is in progress to assess cements prepared after vacuum mixing as well as by hand mixing, with testing at 37C in saline and for a wider range of stresses.
The IRC gratefully acknowledges the support of the EPSRC for its core grant. The authors also thank Dr. E. Dingeldein and Dr. H. Wahlig, Coripharm GmbH, Germany for their assistance in the supply of some of the bone cements tested.

REFERENCES
1. Kotzar GM, Davy DT, Goldberg VM. Telemetrized in vivo hip joint force data: a report on two patients after total hip surgery. J Ortho Res 1991;9:621 633. 2. Gruen TA, McNiece GM, Amstutz HC. Modes of failure of cemented stem-type femoral component. Clin Ortho Rel Res 1979;141:1727. 3. Topoleski LDT, Ducheyne P, Cuckler JM. A fractographic analysis of in vivo poly(methylmethacrylate) bone cement failure mechanisms. J Biomed Mater Res 1990;24:135154.

Figure 9. Fatigue fracture e surface of Boneloc bone cement displaying layers of material.

616

HARPER AND BONFIELD

4. Culleton TP, Prendergast P, Taylor D. Fatigue failure in the cement mantle of an articial hip joint. Clin Mater 1993;12:95 102. 5. Malchau H, Herberts P. Prognosis of total hip replacement and revision rate in THR: A revision risk study of 148,359 primary operations. New Orleans: 65th Ann Meet Am Acad Ortho Surg; 1998. 6. Krause WR, Mathis RS. Fatigue properties of acrylic bone cements: Review of the literature. J Biomed Mater Res Appl Biomater 1988;22:3753. 7. Lewis G. Properties of acrylic bone cement: State of the art review. J Biomed Mater Res Appl Biomater 1997;38:155182. 8. ISO 5833: 1992(E), International Organization for Standardization, Switzerland (1992). 9. Johnson LG. The statistical treatment of fatigue experiments. London: Elsevier; 1964. p 46 47. 10. Rees DG. Essential statistics. 2nd Edition. London: Chapman Hall; 1989; p 123124, 248 249. 11. Lewis G. Effect of mixing method and storage temperature of cement constituents on the fatigue and porosity of acrylic bone cement. J Biomed Mater Res Appl Biomater 1999;48:143149. 12. Bhambri SK, Gilbertson LN. Micromechanisms of fatigue crack initiation and propagation in bone cements. J Biomed Mater Res 1995;29:233237. 13. Topoleski LDT, Ducheyne P, Cuckler JM. Microstructural pathway of fracture in poly(methylmethacrylate) bone cement. Biomater 1993;14:11651172. 14. Tepic S, Soltesz U. Inuence of gamma sterilization on the fatigue strength of bone cement. Atlanta: Proc 42nd Ann Meet Ortho Res Soc; 1996. p 445.

15. Harper EJ, Braden M, Boneld W, Dingeldein E, Wahlig H. Inuence of sterilization upon a range of properties of experimental bone cements. J Mater Sci Mater Med 1997;8:849 853. 16. KindtLarsen T, Smith DB, SteenJensen J. Innovations in acrylic bone cement and application equipment. J Appl Biomater 1995;6:75 83. 17. Crowninshield RD, Brand RA, Johnston RC, Milroy JC. An analysis of femoral component stem design in total hip arthroplasty. J Bone Jt Surg 1980;62:68 78. 18. Kusy RP. Characterisation of self-curing acrylic bone cements. J Biomed Mater Res 1978;12:271305. 19. Weber SC, Bargar WL. A comparison of the mechanical properties of Simplex, Zimmer and Zimmer low viscosity cements. Biomat Med Dev Art Org 1983;11:312. 20. Davies JP, Burke DW, OConner DO, Harris WH. Comparison of the fatigue characteristics of centrifuged and uncentrifuged Simplex P bone cement. J Ortho Res 1987;5:366 371. 21. Gates EI, Carter DR, Harris WH. Comparative fatigue behaviour of different bone cements. Clin Ortho Rel Res 1984;189: 294 299. 22. Krause W, Mathis RS, Grimes L. Fatigue properties of acrylic bone cement: S-N, P-N and P-S-N data. J Biomed Mater Res Appl Biomater 1988;22:221224. 23. Davies JP, OConner DO, Burke DW, Harris WH. Inuence of antibiotic impregnation on the fatigue life of Simplex P and Palacos R acrylic bone cements, with and without centrifugation. J Biomed Mater Res 1989;23:379 397. 24. Thanner J, FreijLarsson C, Ka rrholm J, Malchau H, Wessle n B. Evaluation of Boneloc. Acta Ortho Scand 1995;66:207214.