PUBLISHABLE EXECUTIVESUMMARY Description of project objectives The main objective of the project is the development of an appropriate Metal Injection Moulding (MIM) technology for Titanium processing, employing a novel plasma atomised powder of perfect sphericity for manufacturing high quality biomedical implant devices, meeting the stringent standards for this kind of materials in a cost-effective way. The specific objectives that should have been met are as follows:
To define the desired properties of dental implants based on international literature and medical standards. Also the design of the elements for dental implants prototype had to be reviewed based on the design principles of reference implant materials reported in literature. A proposal of the design of two dental implants based on biomechanical criteria, limitation of the sintering technique and evaluation of the functional stresses by finite element analysis had to be made.
To perform a comparative mechanical, chemical and surface characterizations of the dental implant prototypes in relation to two commercial dental implants.
To evaluate the biological performance of the dental implant prototypes.
To develop the mixing of the feedstocks and their characterisation in order to investigate the processing parameters in a laboratory scale.
To compare different mixing methods.
To design the prototypes and their moulds. Preliminary experiments with Titanium powder should be made to define the limitations of this material in production.
To optimize the debinding process.
To continue making core shooting experiments.
To design sintering strategies to achieve functional porosity.
To complete work on metal injection moulding processing, debinding and sintering studies.
To characterise the injected, debound and sintered samples.
Work has consisted in working with two different feedstocks and carrying out modifications to the simple geometry mould to improve the injection conditions. It has been possible with one of the feedstocks (Mix3-b1) to obtain parts without defects, as ascertained in the characterisation stage, and with acceptable 30 seconds long injection cycles, which is a major improvement with respect to the previous situation.
The solvent debinding problems have been solved satisfactorily and the thermal cycle has been studied and modified until defect free brown parts have been obtained.
Finally, the sintering work carried out show that good results can be obtained if elements pick-up is controlled; oxygen entry during thermal debinding being the main problem. This indicates that there is a limit to the thickness of parts that can be processed with this feedstock, about 4 mm, which rules out the simple geometry part.
Good progress has been obtained by solving all the major problems that barred the applicability of the developed feedstock to produce MIM parts, except the oxygen pick-up problem.
Experiments showed that good results are also within reach for the complex geometry with the most interesting application being small typical parts in dental implants.
Contractors involved Table 1. Person months each participant worked in each workpackage: WP1: Definition of materials, process and implants requirements
Person-months per participant:
WP2: Development, optimisation and manufacturing of the required powders
Person-months per participant:
WP3: Processing studies
Person-months per participant:
WP4: Prototype implant design and mould manufacturing
Person-months per participant:
WP5: Implants Industrial Development
Person-months per participant:
WP6: Implants evaluation
Person-months per participant:
WP7: Dissemination and Exploitation of project results
The same for all
Person-months per participant:
WP8: Project coordination and management
Person-months per participant:
Co-ordinator: Dr Michalis Vardavoulias
fax: +30 22920 69202
Work performed per work package (WP) and end results
WP1. Definition of materials, process and implants requirements
The available bibliography, the implants that are already on the market, the limitations of the MIM technology and the 3D finite element analysis of two commercial types of dental implants have been our sources to conclude in the design of new implants, using metal injection moulding technique.
The first pattern concerns of a press-fit type of implant that has cylindrical shape with a tapered apical ending. Its length is 13mm and the cylinder’s diameter is 3.5mm. The fixture has a funnel shape cervical ending.
There are some problems concerning the structure.
The first one is the so-called trans-gingival collar, concerning a zone of 1.5-2mm height that locates at the cervical ending of the fixture. This zone is useful for the proper adherence of the epithelium on the implant and so far has been completely smooth. Recently, implants like NobelReplace Tapered and Certain PREVAIL implant of the 3i company, have trans-gingival collar, with their surface similar to that of the fixture. A treatment similar to that is suggested for the gingival area of MIM implants.
The second problem is the way that the abutment is positioned into the fixture. It is suggested that the abutment must be positioned in an internal octagon and restrained into the fixture by screwing in an internal 3-thread driver.
The third problem is the porous character of the construction, where the external surface (thickness: 1 mm) should be much more porous than the internal.
The second pattern concerns a threaded implant of the same shape and dimensions as the press-fit one, which differs only at the spiral which begins under the trans-gingival collar and extends till the apical tapered ending.
Depatment of Biomaterials of National and Kapodistrian University of Athens used three-dimensional Finite Elements to study the mechanical performance of a specific implant – abutment connection for transverse (45N), compressive (450N) and combined loading using the commercial FEA software package Cosmos Works. The experiment took place under the assumption that the implant is externally fully fixed (all degrees of freedom are set equal to zero). The abutment is well screwed in the implant and it is assumed that tolerance should be about 2%. Also the abutment is free to move independently of the implant. The results revealed that the developed stresses were 300MPa for the transverse load, ~500MPa for the compressive load and ~520MPa for the combined loading.
Moreover the bone – implant interaction and more specifically the stresses induced by the implant on the surrounding bone tissue was investigated by Finite Element Analysis. The maximum von Misses stress was observed in the cortical bone adjacent to the implant neck for both implants. The stresses recorded were about 85 MPa for the Replace-Biocare implant and about 70 MPa for the Straumann implant. Very low stresses were recorded in the cancellous bone. An equivalent von Misses stress of about 200MPa was observed in the three locking corners of the Replace-Biocare implant abutment. The same value was recorded for the Straumann implant at the contact area between the abutment and the fixture.
Von Misses Stress in the bone structure
Based on these results we proposed two designs of implants (below). The first pattern was a press-fit type implant that had cylindrical shape with a tapered apical ending. Its length was 13mm and the cylinder’s diameter was 3.5mm. The fixture has a funnel shape cervical ending. The second pattern concerns of a threaded implant of the same shape and dimensions as the press-fit one, differs from that only at the spiral that begin under the trans-gingival collar and extend till the apical tapered ending.
Cylindrical Intra-osseous dental implant
Threaded Intra-osseous dental implant
Characterisation of the sintered material for the MIM technique should provide information on hardness, microstructure, tensile properties such as UTS, YS and E%, and final level of I.I.’s. Ideal results on the sintered parts should show on the hardness data an inward profile from the sintered sample surface to the Ti alloy matrix. The level of I.I.’s on the sintered parts and their mechanical properties should fall within those specified in tables 2 and 3:
Table 1. Acceptable interstitial impurity levels.
Table 2. Material properties.
1 ISO 5832/II-1978
2 ISO 5832-3-1990
The specifications of the titanium powder to be used during the MIM processing, with specific attention on the composition, purity level, size range and rheological properties are the use of completely spherical particles, oxygen content below 1500 ppm with the optimum target below 1000 ppm, and a particle size of lower than 45 m.
WP2: Development, optimisation and manufacturing of the required powders. The objectives in this package involve two separate aims. Firstly the development, optimisation and manufacturing of the required powders, and secondly the optimisation of the plasma atomisation process in terms of productivity, powder particle size distribution and purity in order to further meet technical demands and reduce the cost of powder production.
The theoretical process simulation of power production and identification of the process parameters is to optimise the processing parameters of the plasma atomisation process, in order to provide the adequate Ti particles (re. size and shape) for the further processes. The computer simulation of the powder process involved a series of steps beginning from the problem definition and identification of technological and process parameters followed by the required programming development. The goal of this task is to identify the technological parameters that will lead to a powder production containing a greater amount of particles within the < 45 m size range.
The technological and process parameters as well as the powder characteristics to be studied are delineated in table 4:
The parameters highlighted in red are termed the optimising factors, those which through their variation an optimum value for the particle size distribution (objective function) will be obtained. The method used to determine the mathematical model of the process is based on experimental study together with statistical method called multiple regression method. Through a variation in the optimising parameters, delineated in table 5, optimal process parameters for the production of the desired particle size have been calculated.
Table 4 Optimising parameters.
x1 = I (A)
x2 = f (mm/min)
Gas flow rate
x3 = Q (l/min)
x4 = U (V)
Based on the above parameters three main research directions were taken, firstly an experimental model determining the connection between technological parameters and powder characteristics. Second, a first mathematical model considering the relationship between technological-process parameters and powder characteristics before the secondary atomisation, studying particle formation. And finally, a second mathematical model determining particle behaviour after the secondary atomization, during the cooling process, studying the formation of the sphere shaped particles. Mathematical models are then further developed based on experimental studies necessary to establish some important values such as dynamic viscosity of the liquid and surface tension of the melted metal drops.
The computer simulations performed concluded that the variable process parameters have a distinct and complex influence on the diameters’ arrays studied, i.e. 45-75 m and less than 45 m. A hierarchy also can be made with the parameters related to their contribution to process changes as follows: current intensity, voltage, gas flow and wire feed. It was further concluded that the best performances occur at low feed rates and greater intensities. These influences are more pronounced for diameters array of 45 - 75 m. In case of smaller intensity values the feed has virtually no influence on the process. It is therefore recommended that the following values for the technological parameters will yield optimum results: I – 250 A; U – 108 V; f – 400…450 mm/min; Q – 90…100 l/min. However, the use of these parameters could lead to exceeding the critical values of the working power and cause premature wear of the equipment.
The Investigation of the key plasma atomisation parameters was based on the study of plasma atomization parameters namely wire composition and feed rate, plasma gas composition, plasma power and reactor environment (pressure and gas atmosphere), and the effect these have over powder particle size distribution as well as powder chemical composition.
The current, plasma torch voltage, Ti wire feed rate and gas flow were varied. The particles hence produced were sieved into four size distributions and the results communicated to TUIASI. Through an iteration process between production (Pyrogenesis) and simulation (TUIASI) the ideal parameters to increase -45m diameter spherical particles of Ti were obtained as follows:
Table 5 Ideal parameters.
Process current 245 A
Voltage 108 V
Plasma gas flow rate 90 slpm
Wire diameter 3.2 mm
Wire feed rate 760 mm/min
Optimisation of the design of the atomisation system of the torch nozzles for powder atomisation and also design and incorporation of a wire preheating induction system was used to increase the atomisation. Alternative designs have been examined for the gun nozzles (geometry and construction material). The gun nozzles are currently manufactured from graphite, resulting in reduced time life and powder pollution with carbon traces. Further, the increase of the process productivity by the incorporation in the reaction system of a wire preheating system has been successfully performed. Up to now the material to be atomized is in a wire form without any previous pretreatment before its introduction in the reactor. It is believed that the preheating of the material would increase the wire feed rate, enhancing by this way the atomization rate and powder productivity. For this purpose a wire preheating induction system has been designed and manufactured.
In order to increase the life of the nozzle and the production cycle tungsten nozzles were produced. The geometry of the nozzle was changed into a Laval nozzle in order to increase the plasma gas speed. The result was an approximately 15% increase of the nozzle life. However, no effect to the powder purity was noticed. A prototype preheating system was also investigated where the Ti wire was heated before entering the plasma atomisation chamber. This system is now fully functional at the Pyrogenesis facilities. The system is based on inductive heating with the possibilities of reaching a temperature of up to ~ 800ºC. With the new setup, the metal wire is passed through the heater before atomisation. The preheated wire is fed into the common apex of the three plasma torches, where it is melted. The molten metal is dispersed into a large number of small droplets. As the droplets fall into the chamber, solidification and spherodisation of the Ti powder occurs. The final powder exhibits better particle size distribution than through the previous method, with an increase in the fine particle size. Moreover, the wire feed rate has been increased from 760 mm/min to 1100 mm/min, indicating a corresponding increase in the productivity of the system. However, initial results showed that there was an increase in the oxygen content of the powder to < 0.3 % from < 0.15 %. For this reason, a complementary Ar inlet was desgined, allowing Ar to enter the preheating system, creating an inert atmosphere and eliminating the reaction of Ti with oxygen. After implementation of this fitting, no other alterations were required, since this modification aimed only at reducing the oxygen content, without affecting any other properties. Tables 7 and 8 illustrate the above results:
Table 6 Composition of Ti powder.
Powder - 45 μm
Preheated Powder - 45 μm
Preheated Powder - 45 μm + Ar
Table 7 Particle size distribution.
Particle Size (μm)
With preheating and Ar
- 75 + 45
- 125 + 75
Further work in this task will be done by performing the empirical and theoretical modelling of the system with inclusion of the pre-heating system and the Ar inlet. Initial studies have been performed by TUIASI where by the analysis numerically simulates the behaviour of a 1000 mm length titanium wire which is heated by 3 plasma torches in the Plasma Atomization process. The plasma torches were represented by truncated cones. The analysis was done considering the following titanium properties:
Table 8 Titanium properties.
Thermal expansion coefficient
The analysis was done with two different variants as follows:
A steady thermal analysis, where the wire is considered to be in contact with 3 heat sources at a temperature of 3000˚C and which transmits the heat to the wire through film coefficients (conduction and convection transfer of heat).
A transient thermal analysis, where the wire is considered to be in the proximity of the heat sources but not in direct contact (radiation transfer of heat appears, also).
A new nickel-free shape memory titanium alloy powder has been developed by Erothitan (Germany). Within this task the shape memory powder will be optimised and produced according to the specifications defined in workpackage 1.
WP3. Processing studies Using the powder of Pyrogenesis and that of Erothitan IFAM produced feedstocks and characterised them with respect to their rheology and homogeneity. We found that the moulding of the feedstock was possible and the homogeneity was good. Some problems occur if a pre-mixed binder is used as was done in the preliminary experiments.
Overall, the simple geometry mould has been modified and injection, debinding, sintering and characterisation work has been carried out with two different feedstocks. The overall result has been to identify a feedstock composition and MIM processing conditions that will allow the fabrication of small parts, such as the ones required, typically, for the industrial application envisaged in the project, although the only main problem still remaining - excessive oxygen pick-up during thermal debinding. An analysis of the two powders used in this project was performed and debinding of the fine fractions of these powder tested in air. This test resulted in the decision to concentrate on argon debinding for the powder fraction of -45 µm. Small amounts of Biotinol powder were analysed similarly. Based on the binder developed, the process of mixing feedstocks and their homogeneity ware investigated. Some materials were provided for testing their mixing and characterization potential. Problems arose with moulding the feedstock were solved by remixing the feedstocks resulting in improved processing. The feedstocks were analysed in order to simulate the mould filling during MIM. Based on this data mould filling simulation was performed using the software SIGMAsoft. This showed that filling of the component should present no problems using the projected feedstock. Nevertheless, some uncertainty remains as some effects like jetting and phase separation are not taken into account. The feedstock was also analysed with respect to its phase separation. It was found that phase separation occurs but the importance cannot be foreseen. Experiments were performed to do Rapid Prototyping using titanium powders in order to produce some prototypes. These were the basis to actually produce prototypes of the dental demonstrators using 3D data. Detailed improvements to the mould designs were made in order to achieve a design suitable for MIM.
Using the plasma atomization technology production of Ti particles was achieved. The work was based on the computer simulation of the process parameters. Computer and mathematical models were developed for the plasma process, as detailed in the report with main target to produce smaller particles. The suggestion to achieve this was to preheat the metal before feeding.
An extension period was required in order to solve some major issues with regards to the injection moulding of the titanium powders and subsequent issues such as debinding and mould making.
Left hand side picture: The part shown to the left has been processed lying on its axis on a powdered bed and shows no deformation, whereas the part on the right hand side was processed standing on its lower end and is deformed. Right hand side picture: Detail of the undeformed part showing a uniform, defect-free microstructure.
Results of both SEM and porosity show that there are differences with the choice of binder. Agar binder retains the greatest porosity and allows little degree of sintering within particles. Cellosize is the worst binder, allowing almost full sintering within the particles. Acrawax and Polyox are average binders.
Mechanical testing of the samples were performed in order to correlate strength with porosity and be able to obtain the ideal candidate. Definitely -125 mm sized particles are preferable to -45 mm particles in order to retain porosity. Composites Ti-HA with acceptable pore structures can be also obtained using the materials prepared in task 3.5.
The SLA surface of a Standard Plus Implant of Straumman company and Osseospeed surface of an Astra Tech implant of Astra company were analysed. Analysis included Scanning Electron Microscopy (SEM) with SEI images and Atomic Force Microscopy (AFM).
SEM analysis of commercial dental implants (Standard Plus Implant) revealed that the sandblasting procedure leads to the formation of impressions which size is about 50 μm. This size is fairly the same with the size of osteoblasts and that leads to a better adhesion of the osteblasts on the implant surface (Fig. 1).
From AFM analysis in a commercial dental implant (Astra tech osseospeed) it is revealed that the surface morphology at nanoscale level allows the punctual deposition of the nano biological molecules between the hills and the valleys of the surface topography (Fig. 2).
The different treatments affect both implant surface topography and composition, a fact which has a direct influence on the biologic events during osseointegration.
It is now well documented that treated rough implant surfaces lead to more bone apposition, enhanced interfacial strength and reduced healing time at the bone-implant interface. Also, the smooth implant surface topography influences epithelial and connective cell attachment, orientation and movement at the soft tissue-implant interface.
From the results of the present study it is concluded that the roughness of the surface of the pursuing implant must wave in a nanoscale level in order to allow the biomolecules to attach on it.
WP4: Prototype implant design and mould manufacturing Responsible partner: TEHNOTON. Partners involved: FTS, DENA, Pyrogenesis, Erothitan, DTS, NKUA, IFAM and INASMET. Main objectives were to define all details of the dental implants to be developed by MIM, to design the prototype implant shape and to design a prototype with minimal installation and functional stresses to the supporting tissues. Finally it involved the manufacture of the moulds of one prototype spinal and one dental implant. Nevertheless, the objectives have been slightly modified (as previously mentioned) to the design and manufacture of only one dental implant prototype which involved the manufacture of two mould components, thereby still lying within the initial project specifications.
Tehnoton has designed the MIM moulds and also has made the drawings for the implant. Also the technical details for the mould manufacturing have been established in order to obtain a first mould.
Mould design by partner Tehnoton Tehnoton designed the MIM mould after discussion with IFAM. Some disadvantages were found in the design. The design was good for plastic injection moulding but did not take into consideration the special needs of MIM. Some critical points were:
Large surface area on the injection side, components will stick in the injection side and the gate will break due to the low strength of the feedstock when the sprue sticks to the ejection side
No possibility for venting on the injection side which is the end of the filling cycle
IFAM suggested to set the part half into both mould halves and use a core to create the hole. The gate should be at the tip of the components, moulding from thick to thin. Both halves should be cooled. The ejectors need to be contoured and fairly large. A more detailed discussion will be necessary with partner TiJet and their mould producer.
The next version of the mould was also discussed. Tehnoton had designed a very complicated mould in order to incorporate all the details of the design. Based on the discussion between the partners with knowledge on MIM and mould making and taking into account the needed functionality of the implant a slightly simpler implant design was chosen. Tehnoton then redesigned the mould and managed to make it much simpler. Now, the industrial production of the dental implants has become feasible.
Tehnoton has started by defining the critical steps towards the design of moulds.