Medical Engineering & Physics 20 (1998) 620–624

Technical Note

Hip joint prosthesis design: effect of stem introducers

K.J. Mathiasa,*, J.C. Leahya, A. Heatona, W.F. Deansb, D.W.L. Hukins a

a Department of Bio-medical Physics and Bio-Engineering, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, UKb Department of Engineering, University of Aberdeen, Foresterhill, Aberdeen AB24 3UE, UK

Received 6 February 1998; received in revised form 21 August 1998; accepted 21 August 1998

Abstract

The stress levels in the femoral component of a total hip prosthesis (Corin Taper Fit, Corin Medical Ltd, Cirencester, Gloucester-shire, UK) were calculated by finite element (FE) analysis. This prosthesis has two holes drilled in the shoulder to engage a stemintroducer. There were no unacceptable stress levels around these holes. Instead the maximum stresses were around the peripheryof the shaft of the stem, as has been observed for FE analyses of conventional designs. Three prostheses were also subjected tocyclic mechanical testing (peak load 2.3 kN) according to the appropriate British Standard. The holes were examined for cracks,before and after testing, by stereomicroscopy. All three specimens were able to withstand 5 million loading cycles with no evidenceof damage. Thus it is possible to design a femoral component with holes in the shoulder, to accommodate a stem introducer, withoutcreating unacceptable stress concentrations. 1998 IPEM. Published by Elsevier Science Ltd. All rights reserved.

Keywords:Finite element modelling; Stem introducer; Hip joint prosthesis; Mechanical testing

1. Introduction

Recent designs for total hip replacement prosthesesmay incorporate features which enable an introducer tobe used to position the femoral component. Somedesigns of stem introducer simply grasp the neck of thefemoral component, but others incorporate a rod whichfits into a hole drilled in the shoulder of the prosthesis.For example, the femoral component of the Corin TaperFit prosthesis (Corin Medical Ltd, Gloucestershire, UK)incorporates two holes which are drilled into itsshoulder. The introducer then engages these holes forinsertion of the prosthesis; it is designed to disengagewhen the prosthesis is secured in its final position.

Some surgeons have expressed the concern that suchfeatures may weaken the stem by introducing unaccept-able stress concentrations. We have investigated thisconcern, for the specific case of the Corin Taper Fit pros-thesis, by finite element (FE) modelling and by cycliccompressive loading. The combination of FE analysisand experiment is considered to be “many times more

* Corresponding author. Tel.:1 44-(0)1224-681818 ext. 52433;Fax: 1 44-(0)1224-685645; E-mail: k.mathias@biomed.abdn.ac.uk

1350-4533/98/$ - see front matter 1998 IPEM. Published by Elsevier Science Ltd. All rights reserved.PII: S1350-4533 (98)00065-4

Fig. 1. Three-dimensional FE model of the prosthesis (narrow stemdesign). This figure shows the coordinate system used to describe thetilting of the stem.

621K.J. Mathias et al. /Medical Engineering & Physics 20 (1998) 620–624

Fig. 2. Meshed three-dimensional model of the prosthesis. Note the small mesh size in the shoulder, allowing the region around the holes to beexamined in detail. The load was applied vertically (in the negativey-axis direction) through the centre of the head and arrows indicate the regionwhere the outer surface of the stem was constrained.

powerful than the sum of their individual applications”[1].

2. Materials and methods

2.1. Finite element analysis

FE analysis was performed on both versions of thestem which are characterised by offsets of 38 mm(‘narrow stem’) and 45 mm (‘standard stem’). Modelswere generated, analysed and post processed using theANSYS finite element package (version 5.1; ANSYSInc., Houston, PA, USA). They were based on drawingssupplied by Corin Medical Ltd: Issue No. 4 of 30October 1995 for the narrow stem and Issue No. 4 of 1November 1995 for the standard stem. It was possibleto analyse a two-dimensional model of the standard stemwith plane strain but with a finite thickness. The upperpart of the model was assigned a thickness of 11 mm;in the real prosthesis it tapers from 12 to 9.4 mm. Thelower part was assigned a thickness of 6 mm, corre-sponding to a taper of 9.4 to 4 mm. Holes were assigneda diameter of 6 mm. The model was described as a mesh

of eight-node quadratic elements which are well suited tomodelling curved boundaries [2]. This two-dimensionalmodel gave satisfactory results for the standard stem, butas the holes in the narrow stem lie close to the curvedshoulder, it was necessary to build a full three-dimen-sional model for it, shown in Fig. 1. This is the modelwhich will be discussed in this paper. In any case, it wasdesirable to produce as realistic model as possible forthe narrow stem because the dimensions and positionsof the holes were more likely to result in unacceptablestress concentrations than for the standard stem. It wasmodelled as a mesh of three-dimensional ten-node tetra-hedral structural solid elements which are well suited tosuch irregular shapes [2].

Fig. 2 shows the meshed and constrained three-dimen-sional model of the femoral component with a narrowstem and defines anx-axis (horizontal) and ay-axis(vertical); the z-axis is perpendicular tox and y. Theelement dimensions were varied to obtain a fine mesharound the points of interest. The whole model was thenvisually checked for any abnormal element aspect ratios.Models were tilted about thex-axis by 9°, to introducea torsional component to the loading, and about thez-axis by 10°, to mimic its insertion into the femur, to

622 K.J. Mathias et al. /Medical Engineering & Physics 20 (1998) 620–624

Fig. 3. Stress levels in the stem of the prosthesis, with the highest stresses at the top of the restrained nodes, where the bone cement would finish(three-dimensional FE model). The stresses are von Mises and quoted in N/mm2 (1 MPa).

conform with the British Standard for testing total hipprostheses (BS 7251: Part 5: 1990). The Young’s modu-lus of the steel stem (Sandvik Bioline REX 734; SandvikSteel, Sandviken, Sweden) was 200 GPa [3] and Pois-son’s ratio 0.30 [4]. A static point load of 2.5 kN wasapplied vertically through the centre of the head of thefemoral component. This load is approximately threetimes body weight, which is of the order of the loadapplied to a prosthesis in the living body [5]. All externalnodes in the model below 80 mm from the point of loadapplication were constrained to have a displacement ofzero to comply with the fixation of the stem in the BritishStandard for testing total hip prostheses (BS 7251: Part5: 1990). This was not meant to mimic the situation invivo, but was used in order to compare the results of theFE analysis with those obtained in mechanical testing.

2.2. Mechanical testing

Three femoral components of the Corin Taper Fitprosthesis were investigated. All test specimens con-sisted of a narrow stem with a 28 mm diameter femoralhead. They were examined for cracks, both before and

after mechanical testing, by variable magnification stere-omicroscopy. Any debris in the holes was removedbefore examination. During examination, holes werealigned both perpendicular and at an angle to the opticalpath. Each hole was examined from the base to the stemsurface (at a magnification of 8–10 times) and any sur-face features examined at higher magnification.

The three specimens were subjected to cyclic mechan-ical testing according to the appropriate British Stan-dards (BS 7251: Part 5: 1990 and BS 7251: Part 12:1995). They were embedded in Technovit 3040; (LecoInstruments UK Ltd, Stockport, Cheshire) so that theywere maintained with comparable fixation and tilts asdescribed for the FE analysis. The femoral head articu-lated within a ultra high molecular weight polyethyleneblock (UHMWPe, Nylonic Engineering, Cumbernauld,Scotland) mounted on a hardened steel plate. Com-pression was applied vertically through a roller bearingto minimise forces in the horizontal direction. Specimenswere surrounded by physiological saline, maintained at37°C, which was continuously aerated throughout thetest. They were sinusoidally loaded between 0.3 and 2.3kN at a frequency of 5 Hz for 5 million cycles (as speci-

623K.J. Mathias et al. /Medical Engineering & Physics 20 (1998) 620–624

Fig. 4. Stress levels in the region of interest of the prosthesis, looking down on the holes, showing no regions of high stress. The stresses arevon Mises and quoted in N/mm2 (1 MPa).

fied in BS 7251) using an Instron 8511 mechanical test-ing machine (Instron Ltd, High Wycombe, Bucks), i.e.the duration of each test was 11.6 days.

3. Results

The FE models show that the maximum stress isaround the periphery of the shaft of the stem and not inthe region of the holes which are intended to accept thestem introducer. Therefore, there are no unacceptablestress concentrations around these holes. Fig. 3 showsthe stress distribution calculated from the three-dimen-sional model for the narrow design. As expected thehighest stresses occur at the top of the restrained nodes,due to the induced bending moment. All stress arereported as von Mises stresses. If these restrained nodesare removed, as shown in Fig. 4, then the stresses at thepoint of interest can be seen. The maximum stressaround the holes was in the range 160–200 MPa, whichis much lower than the range of 320–360 MPa aroundthe periphery of the stem. For the standard design, thestress in a small region around the medial hole was inthe range 140–190 MPa, but for most regions around the

holes it did not exceed 90–160 MPa. However, aroundthe periphery of the stem the stresses were in the regionof 380–430 MPa.

All specimens survived 5 million cycles of mechanicaltesting with no evidence of failure. Examination of theholes, by stereomicroscopy, showed no evidence ofcrack initiation.

4. Discussion

FE models showed that the two holes in the stem ofthe Corin Taper Fit prosthesis do not lead to unaccept-able stress concentrations. The stress around the periph-ery of the stem shaft was at least twice as high as thatin the shoulder region where the stem is located. Thisresult is consistent with FE analyses of prosthesescemented into the femur which indicate that the stressis greatest in the middle of the shaft [6,7]. The stress inthe shaft is believed to be acceptable for two reasons.Firstly, this part of the stem closely resembles conven-tional designs (e.g. the Exeter hip, Howmedica, London)which have proved successful in the past. Secondly, thefatigue limit for stainless steel in bending is approxi-

624 K.J. Mathias et al. /Medical Engineering & Physics 20 (1998) 620–624

mately 460 MPa [8] for a maximum tensile strength of870 MPa. This figure depends on parameters such assurface finish and the method of machining. This fatiguelimit continues to increase for steels with tensilestrengths of up to 1100 MPa and then starts to level out.As the tensile strength of the steel used for the prosthesisis 1460 MPa, the stresses encountered here will be belowthis fatigue level. Since the stresses in the shaft areacceptable, the lower stresses in the shoulder region,around the holes, must be acceptable too.

The predictions of the FE analysis were consistentwith the mechanical testing. Tests were performed onthe smaller of the two femoral components manufactured(38 mm). Both the FE analysis and simple mechanicalconsiderations predict that this is the size which is mostlikely to have unacceptable stress concentrations aroundthe holes in the shoulder. However, specimens were ableto withstand 5 million applications of a compressive loadcorresponding to about three times body weight.

The general principle which emerges from theseresults is that it is possible to design a femoral compo-nent for a total hip prosthesis which has holes in theshoulder, to accept an introducer, without creating unac-ceptable stress concentrations at these holes. However,it is important to note that the results were obtained fromone design of prosthesis (the Corin Taper Fit). Futuredesigns will also be needed to be subjected to FE analy-sis and a testing procedure, which conforms to some

generally accepted standard, to ensure that they are equ-ally safe.

Acknowledgements

We thank Mr G.P. Ashcroft FRCS for discussion. Fin-ancial support was provided by Corin Medical Ltd andSHEFC (Joint Equipment Initiative Grant).

References

[1] Huiskes R, Hollister SJ. From structure to process, from organ tocell: recent developments of FE-analysis in orthopaedic biomech-anics. J Biomech Eng 1993;115:520–7.

[2] NAFEMS (National Agency for Finite Element Methods andStandards). A finite element primer. UK: NAFEMS Ltd, 1991.

[3] Sandvik Steel. Sandvik Bioline REX 734, Data sheet S-262-ENG.Sandviken, Sweden: Sandvik Steel 1991.

[4] Gere JM, Timoshenko SP. Mechanics of materials. 2nd SI ed.USA: Wadsworth International, 1985:744.

[5] Rydell N. Biomechanics of the hip joint. Clin Orthop1973;92:6–15.

[6] Crowninshield RD, Brand RA, Johnston RC, Milroy JC. An analy-sis of femoral component stem design in total hip arthroplasty. JBone Joint Surg [Am] 1980;62A:68–78.

[7] Andriacchi TP, Galante JO, Belytschko TB, Hampton S. A stressanalysis of the femoral stem in total hip prostheses. J Bone JointSurg [Am] 1976;58A:618–24.

[8] Frost NE, Marsh KJ, Pook LP. Metal fatigue. 1st ed. UK: OxfordUniversity Press, 1974:40–122.