1 ©2010 ISC Research Symposium

Proceedings of the 4th Annual ISC Research Symposium ISCRS 2010

April 21, 2010, Rolla, Missouri

SELECTIVE LASER SINTERING OF 13-93 BIOACTIVE GLASS BONE SCAFFOLDS

Krishna C. R. Kolan1, Ming C. Leu1, Gregory E. Hilmas2, Mariano Velez3 1Department of Mechanical and Aerospace Engineering, Missouri University of Science and Technology,

Rolla, MO 65409, USA 2Department of Materials Science and Engineering, Missouri University of Science and Technology,

Rolla, MO 65409, USA 3Mo-Sci Corporation, Rolla, MO 65401, USA

ABSTRACT Bioactive glasses are promising materials for bone scaffolds due to their ability to assist in tissue regeneration. When implanted in vivo, bioactive glasses can convert to hydroxyapatite, the main mineral constituent of human bone and form a strong bond with the surrounding tissues thus providing an advantage over polymer scaffold materials. Bone scaffold fabrication using additive manufacturing techniques like selective laser sintering (SLS) provides control over design and fabrication of pores in the scaffold which helps in mimicking the real human trabecular/cancellous bone. 13-93 bioactive glass, a third-generation bioactive and resorbable material designed to accelerate the body’s natural ability to heal itself, was used in the research described herein to fabricate bone scaffolds using SLS process. 13-93 bioactive glass mixed with stearic acid (as the polymer binder) by ball milling was used as the powder feedstock for the SLS machine. The fabricated green scaffolds underwent binder burnout to remove the stearic acid binder and then sintered at temperatures between 6750C and 7000C. The sintered scaffolds had pore size varying from 300 µm to 800 µm with 50% apparent porosity and a maximum compressive strength of 24 MPa, among the highest reported for controlled porosity scaffold fabricated with bioactive glasses using SLS process. The scanning electron microscope (SEM) images of the sintered scaffolds shows a rough surface and micro pores (≤ 100 µm) in the walls throw in together with the oriented internal architecture could be encouraging for tissue engineering applications. The X-ray diffraction (XRD) analysis on the sintered scaffolds confirmed the chemical composition of the 13-93 bioactive glass with no signs of crystallization. 1. INTRODUCTION The conventional fabrication of bone tissue scaffolds have relied on traditional materials technique such as modeling and machining to create the shape of the scaffold. These techniques have difficulty providing a tight fit to complex defect sites so healthy bone is usually machined at the time of implantation to make the defect site fit the graft. Particulate

packing had some success but is limited to non-load bearing applications. Rapid prototyping techniques can be used to tailor make synthetic bone tissue scaffolds to the specific geometric requirements of a particular bone defect site. Also, complex internal architectures of scaffolds besides shape, pore size, and porosity can be achieved through solid freeform fabrication methods. A bio-degradable, photocurable copolymer, acrylate-endcapped poly (e-caprolactone-co-trimethylene carbonate) for use in SLA fabrication was developed by Matsuda and Mizutani [1]. An extrusion freeform technique has also been adapted for fabrication of ceramic scaffolds. Hydroxyapatite suspensions were extruded through a range of nozzles and the latticework scaffolds were sintered [2]. Robocasting technique was developed by Cesarano to fabricate HA latticework structure for load-bearing scaffolds [3]. Fused deposition manufacturing has been used to fabricate precision scaffolds of PLA for cartilage and PCL for bone [4]. Fused deposition modeling has been used to fabricate bioresorbable 3D scaffolds with interconnected pore network by Hutmacher [5]. 3D printing process was used to fabricate 3D scaffolds for bone tissue enginnering [6].

Selective laser sintering has been used in fabricating porous scaffolds for over a decade. Several biopolymers were used in fabricating scaffolds using this process and in some cases blended with hydroxyapatite to make them bioactive. Polycaprolactone (PCL) has been used as a prime bioresorbable polymer and also mixed with Hydroxyapatite in fabricating the scaffolds [7-9]. Polyetheretherketone and hydroxyapatite biocomposite blends were also used in fabricating scaffolds [10, 11]. A comprehensive study on different biodegradable polymers used by researchers was done by Thomson et al in [12]. Lorrison and Goodridge have used bioactive glass-ceramics to fabricate scaffolds using this process [13, 14], demonstrating the potential for using SLS to fabricate bioactive glasses. 13-93 bioactive glass has been researched previously by Roger Brown and Qiang Fu in preparing fiber scaffolds and porous scaffolds using polymer foam replication technique respectively [15, 16]. The results show a potential application of 13-93 bioactive glass in

2 ©2010 ISC Research Symposium

repairing bone defects and tissue engineering. Bioactive glasses rapidly (~ 2 h) forms biologically active hydroxyapatite (HA) when their surface is exposed to body fluids, and particles 74 to 105 µm completely convert to HA within a few weeks [17]. They have also been shown to not only be osteoconductive but osteoinductive. The aim of our work was to fabricate a 13-93 bioactive glass bone scaffold using SLS process.

The basic principle of SLS is to coat fine ceramic powders with thermoplastics, which melt when a computer controlled laser beam scans to heat the powders. The thermoplastic acts like a binder to fuse the particles to form a single solid layer. After one layer is finished, according to the computer generated slice pattern from a CAD model, the second layer begins, thereby fabricating a 3D part. A schematic diagram showing inside chamber of the SLS machine is shown in Fig. 1.

2. MATERIALS AND METHODS 2.1. Preparation of powder

13-93 bioactive glass (prepared at Mo-Sci Corp. code: GL-0811) with nominal chemical composition of 53% SiO2, 4% P2O5, 20% CaO, 5% MgO, 6% Na2O and 12% K2O (weight %) was made from high-purity chemical reagents by melting several batches in 5-kg capacity platinum crucibles and quenching in water. The chemical composition of the quenched glass was confirmed by XRF (X-ray fluorescence analysis). The quenched glass was milled and sieved to a particle size below 75 µm. The milled 13-93 bioactive glass was mixed in a V-blender with stearic acid (C18H36O2, grade HS, Acros Organics) and dry ball-milled for 8 h with ZrO2 grinding medium to obtain powders with mixing ratio 50:50 and 60:40 (13-93 glass to stearic acid) by volume.

2.2. Selective laser sintering All the experiments related to fabrication were carried out

on a commercially available DTM Sinterstation 2000 SLS machine at Missouri University of Science and Technology, Rolla. A detail description of the machine and its parameters are explained in the reference [18, 19]. Two different types of laser sintering, direct and indirect have been followed during the initial phase study to establish a feasible set of parameters which could be conducive in fabricating a scaffold using 13-93 bioactive glass. In the direct selective laser sintering, bare 13-93 glass powder is used as feedstock to the SLS machine and in the indirect selective laser sintering, a polymeric binder is mixed with the actual powder to help fuse the powder particles during fabrication. A post processing step is involved in the indirect process to burn out the binder and sinter the bioglass powder particles. Initial set of experiments were conducted in both the processes to understand the behavior of the materials under laser beam. In the direct SLS process, mono layers were fabricated at different laser powers, scan speed and scan spacing to study the effect of parameters. Since the transition temperature of 13-93 bioglass is around 620oC, the parameters like part bed temperature and part heater temperature become unimportant, which otherwise are maintained just below the melting point of binder [20]. It also requires high amount of heat to fuse the bare 13-93 bioglass particles. Laser power was varied from 5W to 50W, scan speed from 2 in/s to 15 in/s exploring different scan spacings. Several mono layers were fabricated and the effect of parameters was investigated. In indirect SLS process, feasible set of parameters are obtained based on the amount of melting of stearic acid and once established, cylindrical scaffolds with required dimensions are fabricated and then post processed. 2.3. Post processing

All the green parts were post processed in a three stage programmable air furnace (Vulcan Benchtop Furnace, Materials Research Center at Missouri S&T). The SEM image analysis was performed on the green part and sintered part to study the effect of heat treatment schedule. The SEM images of the powder samples were acquired using Hitachi S-570 SEM (McNutt Hall at Missouri S&T) and Hitachi S-4700 FESEM (Materials Research Center at Missouri S&T) was used to capture SEM images of fractured surface of the green part and fracture surface of the sintered part. Mechanical testing was done on the sintered scaffolds to determine the compressive strength using Instron 4469 UTM (Materials Testing Laboratory at Missouri S&T). Six samples in each set, sintered at different soaking conditions, were tested with each sample measuring around 16 mm in length and 8 mm in diameter. All the tests were carried out at a cross head speed of 0.5 mm/min. The chemical composition of the sintered scaffolds was confirmed by running an XRD analysis using Philips X-Pert (Materials Research Center at Missouri S&T) on the fractured scaffold.

Figure 1. Schematic diagram of SLS process

3 ©2010 ISC Research Symposium

(a) (b) (d)

3. RESULTS 3.1. Direct selective laser sintering

As the 13-93 bioglass started to soften at laser powers greater than 30W with 0.009 in scan spacing, scan speeds and spacing were varied starting with the lowest possible limit of the SLS machine. The 13-93 bioglass required higher energy densities which could be obtained only by reducing the scan speed and spacing to the minimum on the commercial machine. By varying one parameter and keeping the other two constant, the effect of each parameter is studied and the best results were observed at the scan speed of 2 in/s and 0.003 in scan spacing (minimum scan speed available for the SLS DTM2000 machine). When the laser power is varied from 15W to 25W, best mono layer was obtained at 25W laser power. Laser power greater than 25W had no substantial improvement in the mono layers. A visual inspection of the mono layers fabricated at different set of parameters indicated the bubbling phenomenon [21] which was also reported by other researchers [13]. Figure 2 shows one such mono layer sintered at 20W. The mono layer instead of being a continuous sheet had oriented porosity in the beam scanning direction. The scan direction of the laser can be clearly observed in a single layer with the horizontal walls running through measuring in width which is around the same as the laser spot size. Because of this phenomenon, there was poor bonding between the successive layers while fabricating a 3D part and it was difficult to remove the 3D part from the part bed. XRD of laser sintered samples of 1393 glass powder showed no crystallization and phase changes confirming the original composition. 3.2. Indirect selective laser sintering

In the indirect selective laser sintering approach, a polymeric binder was used to mix with the 13-93 bioglass powder in two different proportions. Unsuccessful sintering of 1393 necessitated the use of polymer binder. Stearic acid was successfully used as a binder in our previous study [18] which provides a uniform coating on the glass particles after ball milling and leaves little or no carbon residue during burnout stage. The energy provided by laser melts stearic acid which fuses the 13-93 particles together to form a green part. Green parts were successfully fabricated using powder compositions

with 40% and 50% binder content by volume. The effect of layer thickness was also studied in the process. 3.2.1. Determination of SLS parameters

An analytical approach was used in determining the feasible set of parameters while fabricating the green part instead of applying a design of experiments method. As mentioned earlier in section 2.2, the important parameters considered are laser power, scan speed, scan spacing and energy density. Energy density is an important parameter which relates the other three parameters [22]. Based on our previous experience of using stearic acid as a binder [19], an experiment was conducted to check the degree of melting of stearic acid by varying the energy densities from 0.84 cal/cm2 to 2.74 cal/cm2. By means of visual inspection, energy density of 1 cal/cm2 was decided to be sufficient enough to melt the stearic acid, which helps in fusing the successive layers while fabricating the part layer-by-layer. The part bed temperature was maintained at 600C, just below the melting point of stearic acid [20], which also explains the low energy density. Laser power and scan speed were varied while scan spacing was set to 0.009 in [19], so that the energy is maintained at 1 cal/cm2. Laser power was maintained between 2W and 3W as higher laser powers meant unnecessary build up of heat and temperature gradient in the part bed. Scan speeds were varied from 3 in/s to 12 in/s. The initial layer thickness was decided to be around 0.006 in based on the largest particle size. Parts measuring 1” X 1” X 0.04” (1 mm thick) were fabricated with the above mentioned parameters. Best bonding between the successive layers was achieved with 3W laser power and 12 in/s scan speed. 3.2.2. Effect of binder content and layer thickness

Delamination was observed in the parts fabricated with 40% binder using same set of parameters. The amount of binder to fuse and hold the successive layers was minimum because of which the layer thickness was reduced to 0.004 in. Figure 3(a-b) shows the comparison of green parts fabricated using 50% binder content and different layer thicknesses. The part with 0.004 in layer thickness had higher green strength than the one with 0.006 in layer thickness and looked

Figure 2. Mono layer sintered at 20W by direct SLS

(c)

Figure 3. a – 0.006”, b – 0.004” with 50% binder. c – 50% binder, d – 40% binder with 0.004” layer thickness

4 ©2010 ISC Research Symposium

Figure 7. DSC/TGA curve of stearic acid

Figure 8. DSC/TGA curve of stearic acid

compact when compared with the former one. As the particle size of the 13-93 glass/stearic acid mixture was in the range of 25 – 75 µm, the layer thickness was not reduced to 0.003 in (the minimum possible in DTM 2000 Sinterstation), which could barely fit the largest particle. Figure 3(c-d) shows the comparison of green parts fabricated at 0.004 in layer thickness using different binder contents. The green part with reduced binder content had higher green strength and had better surface finish and looked compact. To reduce the binder content further, the layer thickness had to be reduced to minimum possible in the machine which demands smaller particle size than available.

Figure 4 shows the CAD model of the porous scaffolds with dimensions of 20 mm length and 10 mm diameter. The pore size in the CAD model was designed to be 1 mm with 58.8% apparent porosity. The unit repeatable cube was used to calculate the porosity of the CAD model. Figure 5(a) shows the green part which measured 10.2 mm in diameter and 20.5 mm in length. Figure 5(b) shows the green part fabricated using a CAD model with designed pore size of 0.7 mm and 69.3% designed porosity. It was difficult to remove the unsintered powder from the pores of green part and was unsuccessful considering the laser spot size of around 450 µm, which results in transfer of heat to unsintered powder particles and unwanted bonding to the actual sintered part [23].

Figure 6 shows the SEM images of the as received 13-

93 bioglass powder particles, bioglass mixed with 50% stearic acid and the fractured surface of the solid cylindrical green part. The coating of fine stearic acid particles on irregular 13-93 bioglass particles can be observed in fig 6(b) and fused 13-

93 bioglass particles in the melted pool of stearic acid can be seen in fig 6(c).

3.3. Post processing 3.3.1. Determination of heat treatment schedule

After the fabrication of the green parts, a heat treatment schedule was developed based on the DSC-TGA curves of stearic acid and 50:50 13-93 bioglass/stearic acid powders. The DSC-TGA curves are shown in fig 7. The DSC curve shows valley peaks at around 740C and 2870C, which correspond to the melting point and decomposition temperature of stearic acid respectively. The start of

decomposition of stearic acid is confirmed in DTA/TGA curve of bioglass/stearic acid (not shown here) which starts to fall at around 2400C and continues till around 2900C. The complete decomposition of stearic acid binder is confirmed at around 5200C in TGA curve after which the curve goes flat indicating the sample losing 25% of its weight which corresponds to 50% of binder in volume ratio. The heating rate is kept to the

Time, h

Tem

pera

ture

, deg

ree

C

20 40 60 80

200

400

600

800

Figure 4. a – Isometric view of CAD model; b – repeatable unit

Figure 5. a – green part with unsintered powder in the pores; b – successful green part

(a) (b)

(a) (b)

5 ©2010 ISC Research Symposium

Figure 6. a – as received 13-93 bio active glass powder (<75µm); b – stearic acid coated 13-93 particles; c – fused 13-93 particles

Table 1. Effect of soaking conditions on shrinkage and porosity

Figure 9. Cross sectional view of a sintered scaffold

minimum possible available for the furnace because of presence of higher binder content. Holding temperatures were designed based on the DSC/DTA and TGA curves to aid the slow melting of the binder at 700C and to avoid the scaffold structure collapse. The second hold is provided at 2400C with slight increase in the heating rate to help the decomposition of the binder. The final 1hr hold given at 5000C is to make sure of decomposition of any leftover carbon particles in the scaffold. The heating rate was increased to 20C/min at 5000C which was maintained until the final sintering temperature. The scaffolds were sintered at various temperatures ranging from 6750C to 6950C and at two different time periods of 0.5 hr and 1 hr, to improve the sintering and mechanical properties of the scaffold. Figure 8 shows the sintering schedule for the green parts. The same sintering schedule was successfully implemented even for parts fabricated with 40% binder. 3.3.2. Structural analysis

The shrinkage of sintered part compared to green part fabricated with 50% binder content is more than those with 40% binder content when sintered at similar conditions of 6850C for 0.5 hr. 23.7% shrinkage in length and 17.7% shrinkage in diameter was observed for the 50:50 parts compared to 20% in length and 14.4% in diameter for 60:40 parts. Time period had more effect on shrinkage than the temperature at which parts were sintered. This can be observed in table 1. Similar observations were also made in the measured porosity of the scaffolds. The apparent porosities of

Soaking conditions

Shrinkage (%)

Porosity (%)

6850C - 0.5 hr 20 53.2 6850C - 1 hr 21.7 50.7

6950C - 0.5 hr 20.5 53.5 6950C - 1 hr 22.6 50.3

scaffolds were measured following ASTM C373 standard. The porosities mentioned in the table might be only indicative

because of comparatively larger pores, the water molecules on the outer pores were absorbed by the moistened cloth, resulting in less measured porosity than the actual value. However, the trend of scaffold porosities should be the same with varied sintered conditions. Figure 9 shows the cross sectional view of a cylindrical disc cut from a scaffold. The pore size varied from 300 µm to 800 µm, which could result in better bone growth than having micro pores [24, 25].

3.3.3. Mechanical properties The scaffolds fabricated with 50:50 powder had a

compressive strength ranging from 5 – 11 MPa based on the soaking conditions in sintering schedule. Low strength was achieved when sintered at 6750C. Therefore, the scaffolds fabricated with 60:40 powder were sintered at 6850C and 6950C at two different time periods which had a compressive strength ranging from 13 – 24 MPa. Figure 10 shows the graph with variation of compressive strength at different soaking conditions. A two way balanced ANOVA test showed the effect of increase in temperature to be more predominant than increase in time period when compressive strength was a response variable. Further increase in sintering temperature to 7000C showed crystallization of silicon phase in 13-93 bioglass in XRD analysis.

(a) (b) (c)

6 ©2010 ISC Research Symposium

Figure 10. Effect of soaking conditions on strength

Figure 11. XRD graphs of processed bioglass compared to original 1393 glass.

Figure 12. a – fractured surface with macro and micro pores; b – micro pores; c – outer surface of the scaffold

3.3.4. Microstructure analysis

Figure 11 shows the XRD results of 13-93 bioglass powder when processed under direct and indirect selective laser sintering. XRD results of the scaffold sintered at 7000C for 1 hr showed crystallization with a peak at around 300 which was identified as silicon phase (not shown here). There was no crystallization of 13-93 bioglass when sintered at 6850C for 1 hr or using direct SLS. This can be seen in figure where both direct and indirect sintering trace similar profiles for the processed material when compared to as received 13-93 bioglass powder. Figure 12 shows the SEM images of fractured area and outer surface of a scaffold sintered at 6850C for 1 hr. Interconnected designed pores and unconnected micro pores (less than 100 µm) in scaffold walls can be observed in fig 12(a). The micro pores in the wall are developed because of the higher amount of binder used in the indirect SLS process, which reduce the strength of the scaffold. The presence of micro pores can be reduced to an extent by increasing the sintering temperature, but cannot be eliminated

in SLS process, where the use of a binder is required. Lesser the micro pores in the scaffold walls, greater is the strength of the scaffold.

(a)

(b)

(c)

7 ©2010 ISC Research Symposium

4. CONCLUSIONS The preliminary results indicate that 13-93 bioactive glass can be successfully selective laser sintered without changing its composition while fabricating bone scaffolds for tissue engineering purposes. Stearic acid was successfully used as a binder without affecting the composition of the bioglass in the process of indirect select laser sintering using commercially available DTM 2000 sinterstation. The average compressive strength of the scaffolds was 20.3 MPa with a maximum of 23.6 MPa, which demonstrate the use of scaffolds to mimic the trabecular/cancellous bone of a human bone when designed with a similar internal architecture. 5. ACKNOWLEDGEMENTS This research work was funded by National Science Foundation under contract no: NSF SBIR phase I award 0912019, Mo-Sci Corporation and Intelligent Systems Center at Missouri S&T. The authors are thankful to Clarissa Wisner, Eric Bohannan (Materials Research Center) and Gurpartap Singh (graduate student) for their technical assistance. 6. REFERENCES [1] T. Matsuda and T. Mizutani, “Liquid Acrylate-endcapped Biodegrable Poly(e-caprolactone-co-trimethylene Carbonate) II Computer-aided sterolithographic microarchitectural surface photoconstructs,” Biomaterials 14 323-330 (2002). [2] F.C. Gomes de Sousa and J.R. Evans, “Sintered Hydroxyapatite Latticework for Bone Substitute,” J. Am. Ceram. Soc., 86 [3] 517-519 (2003). [3] J. Cesarano III,* J.G. Dellinger, M.P. Saavedra, and D.D. Gill, “Customization of Load-Bearing Hydroxyapatite Lattice Scaffolds,” Int. J. Appl. Ceram. Technol., 2 [3] 212–220 (2005). [4] Hsu, S.; Yen, H.; Tseng, C.; Cheng, C. & Tsai, C., “Evaluation of the growth of chondrocytes and osteoblasts seeded into precision scaffolds fabricated by fused deposition manufacturing,” J Biomed Mater Res B Appl Biomater, 2006. [5] Dietmar W. Hutmacher, Thorsten Schantz, Iwan Zein, Kee Woei Ng, Swee Hin Teoh, Kim Cheng Tan, “Mechanical properties and cell cultural response of polycaprolactone scaffolds designed and fabricated via fused deposition modeling,” J. Biomed Mater Res A, 2001. [6] Alberto Rainer, Sara Maria Giannitelli, Franca Abbruzzese, Enrico Traversa, Silvia Licoccia and Marcella Trombetta, “Fabrication of bioactive glass–ceramic foams mimicking human bone portions for regenerative medicine,” Acta Biomaterialia Volume 4, Issue 2, 2008, 362-369. [7] Jessica M. Williams, Suman Das, “Bone tissue engineering using polycaprolactone scaffolds fabricated via selective laser sintering,” Biomaterials 26(2005) 4817 – 4827. [8] M. C. Azevedo, “Development and properties of polycaprolactone/hydroxyapatite composite biomaterials,” Journal of materials science: Materials in medicine 14(2003). [9] F. E. Wiria, Chua, “Polycaprolactone/hydroxyapatite for tissue engineering scaffold fabrication via selective laser sintering,” Acta Biomaterials 3 (2007) 1-12.

[10] Tan, Chua, “Scaffold development using selective laser sintering of PEEK-HA biocomposite blends,” Biomaterials 24(2003) 3115-3123. [11] Tan, Chua, “Fabrication and characterization of 3d PEEK HA biocomposite scaffolds using laser sintering,” Journal of engineering medicine vol 219. [12] R. C. Tjompson, “Biodegradable polymer scaffolds to regenerate organs,” Advances in polymer science. vol 122. [13] J.C. Lorrison, R.D. Goodridge, K.W. Dalgarno, and D.J. Wood, “Selective Laser Sintering of Bioactive Glass-Ceramics,” Solid Freeform Fabrication Proceedings, Austin, Texas, 1-8 (2002). [14] Goodridge RD, K W Dalgarno, and D J Wood, “Indirect selective laser sintering of an apatite–mullite glass-ceramic for potential use in bone replacement applications,” J. Engineering in Med H, 220, 2006. [15] Roger F. Brown, Delbert E. Day, Thomas E. Day, Steve Jung, Mohamed N. Rahaman, Qiang Fu, “Growth and differentiation of osteoblastic cells on 13–93 bioactive glass fibers and scaffolds,” Acta Biomaterialia 4 (2008) 387–396. [16] Qiang Fu, Mohamed N. Rahaman, B. Sonny Bal, Roger F. Brown, Delbert E. Day, “Mechanical and in vitro performance of 13–93 bioactive glass scaffolds prepared by a polymer foam replication technique,” Acta Biomaterialia 4 (2008) 1854–1864. [17] Larry L. Hench and Orjan Andersson, “Bioactive glasses,” Larry Hench and June Wilson, eds., “An Introduction to Bioceramics” World Scientific Publishing Col. Ltd., 1993. [18] M. Leu, E.B. Adamek, T. Huang, G.E. Hilmas, F. Dogan, “Freeform Fabrication of Zirconium Diboride Parts Using Selective Laser Sintering,” Proceedings of Solid Freeform Fabrication Symposium, Austin, TX, August 4-6, 2008. [19] S. Pattnaik, M. C. Leu, G. E. Hilmas, F. Dogan, “Investigation of zirconium diboride parts using selective laser sintering,” Thesis (Master’s), 2009. Missouri S&T, Rolla. [20] Gibson, I. and Shi, D., “Material properties and fabrication parameters in selective laser sintering process,” Rapid Prototyping Journal, 1997, 3 (4), 129-136. [21] K M Fan, W L Cheung, I Gibson, “Movement of powder bed material during the selective laser sintering of bisphenol A polycarbonate,” Rapid Prototyping Journal; 2005; 11, 4. [22] Nelson, J.C., “Selective Laser Sintering: A definition of the process and an empirical sintering model,” Dissertation (PhD), 1993. University of Texas, Austin. [23] Yang, S., Leong, K. F., Du, Z. & Chua, C. K., “The design of scaffolds for use in tissue engineering. Part II.,” Rapid prototyping techniques. Tissue Eng. 8, 1–11 (2002). [24] Ciara M. Murphy, Matthew G. Haugh and Fergal J. O'Brien, “The effect of mean pore size on cell attachment, proliferation and migration in collagen–glycosaminoglycan scaffolds for bone tissue engineering,” Biomaterials, Volume 31, Issue 3, 2010, 461-466. [25] Vassilis Karageorgiou and David Kaplan, “Review: Porosity of 3D biomaterial scaffolds and osteogenesis,” Biomaterials, Volume 26, Issue 27, 2005, 5474-5491.